Me
Do you have any questions with regard to my packages, SlackBuild scripts, other scripts, or my documentation? Or do you have a request to make? Please use this space to write down your ideas and I will try to answer. Other readers are of course also allowed to voice their thoughts about what you write.
Keep your posts on topic please. No flamewars, trolling, or other nastiness allowed. This is not meant to be a replacement for LinuxQuestions.org…
If the blog refuses to accept your post, then perhaps you are affected by a bug in the SQLite plugin. Check if your post contains a string of text which is enclosed by the characters ( ). Is there a pipe symbol or a comma inside those round brackets? Try to remove those and re-post your comment.
Note: this feedback page became too big and is slow to load. Therefore I disabled further comments, renamed the page and opened a new Feedback Page for you.
Or continue to my blog articles if you want.
Eric
Thanks for creating this space, and for being such a force for the Slackware community.
Okay, so the good news is that Handbrake 0.9.8 works perfectly. The bad news is it doesn’t contain any features, only bug fixes. And according to the devs, the current trunk won’t be released as the official build until next year – https://forum.handbrake.fr/viewtopic.php?f=12&t=24951. Is there any way I could get you to build the 4883svn nightly (https://trac.handbrake.fr/browser/trunk?rev=4883)?
hi, just noticed the handbreak package in “restricted_slackbuilds” is still version 0.9.5 in 32/64 versions.
Do I get all the features if I get the regular pkg from slackware.com server?
best regards
O… I forgot that I had to place the packages in restricted_slackbuilds due to the lame and faac encoders… I will set that right tonight and update the repositories.
Actually, it is easy for you to build your own package from a SVN trunk checkout. I will stick to the official releases but you can grab the SlackBuild and edit it so that it has these two lines:
VERSION=${VERSION:-r4883}
RELREV=${RELREV:-“”}
Which causes the script to checkout revision 4883 from trunk and build a package for that.
Eric
Thanks Bob. Unfortunately it’s not going to be that simple, because sourceforge.net only has the official releases. It looks like the only way to grab a nightly for a build is to use a subversion repository call, which is failing badly for me when I try it.
Hm… I reread your post, and understand my confusion now. The current .build file contains no references to SVN, so it defaulted to sourceforge.
Well, after a lengthy process of finding all the dependencies necessary for a compile in unRAID, I have a working executable. I’m not 100% of the best way to make it into a package, though. I ran “make install” in the build directory and that has me going for now, but is there a simple way to make the .build file?
Hi Eric, slackpkg won’t blacklist last multilib update
Mon Jul 30 20:12:24 UTC 2012
current/gcc-4.7.1_multilib-x86_64-1fix1_alien.txz: Rebuilt.
current/gcc-g++-4.7.1_multilib-x86_64-1fix1_alien.txz: Rebuilt.
current/gcc-gfortran-4.7.1_multilib-x86_64-1fix1_alien.txz: Rebuilt. Fixed the
64-bit libgfortranbegin.a library which got overwritten by the 32-bit
version. Thanks to hiptobecubic for reporting this.
current/gcc-gnat-4.7.1_multilib-x86_64-1fix1_alien.txz: Rebuilt.
current/gcc-go-4.7.1_multilib-x86_64-1fix1_alien.txz: Rebuilt.
current/gcc-java-4.7.1_multilib-x86_64-1fix1_alien.txz: Rebuilt.
current/gcc-objc-4.7.1_multilib-x86_64-1fix1_alien.txz: Rebuilt.
I had to add:
“[0-9]+fix1_alien” to /etc/slackpkg/blacklist so they don’t get replace by slackware’s native pkg
best regards
Hi cesarion76
I think I may have to rename the packages so that they end on “fix1_1alien” instead of “1fix1_alien” because having to add yet another line to the blacklist file of slackpkg is not an elegant solution. Thanks for mentioning it.
Cheers, Eric
aiden, I am a bit confused as to what you are trying. A “.build” file? Could you not just use the hints I gave 6 comments back and run the edited handbrake.SlackBuild with changed variable definitions? That will checkout a trunk snapshot and build that for you.
Eric
Bob, yes I tried that first, of course. But I got an “invalid scheme” error. So instead I went down the road of manually checking it out and compiling it locally based on Handbrake’s wiki instructions. What I would like to do is make a package out of that binary so I can post it on the unRAID forums for other users to download. I appreciate your patience, because I’m clearly a novice at these things.
I never saw an “invalid scheme” error. Perhaps if you can post a full log of your failed compilation on http://pastebon.com/ then I could have a look at it.
Eric
Hi. Are there any difficulties in building recent qemu-kvm-1.1.1 for slack 13.1 or 0.14 packages just left from old SlackBuild version?
Thank you
Thanks Eric for the kde 4.9.1, it has helped me out a lot.
Hello Erik!
Could you please create a package for virt-manager? Or create a howto how we can install it? I found some howtos on the net for slackware and virt-manager, but I was unable to make it.
Thank you.
Krisz
Hi Krisz
I have been looking at virt-manager, especially for the VNC viewer features, and it is possible that I will create packages and/or write an article about it, when I get some free time.
Eric
Alien, everyone would like to know how is your perfect Slackware installation, from beginning to end, partitioning, filesystem, packages, tweaks and also a screenshot of your computer, this can be an upcoming article.
Hi willian
My own computers are not all that interesting to talk about. They are functional but not shiny. I add only a few packages to a full Slackware install, depending on the needs (my laptop for work has some other stuff than the desktop I share with the family).
Even the background on this laptop is the standard KDE bakground…
Eric
Hi Eric
First of all I wish to thank you for the amazing packages you place and all the contribution you do for slackware.
In your post about LibreOffice, you mentioned somewhere that packages under 13.37 might work for 14.0.
I wonder if I should try it for lame and others, in the restricted packages, or wine, in the regular ones, and even for packages for older versions, like xawtv for 12.0 (I have multilib installed).
I am using slackware 14.0.
cheers!
Hi Duodecimo.
The golden rule for binary packages is, if they are not available for the Slackware release you are currently using, try a package for an older release. Often that will just work (but not always). If I find a package for Slackware 13.37 which fails on Slackware 14 then I will specifically compile a new package for Slackware 14.
So, yes, you can use the “older” packages for lame and wine. If you encounter any issues, like library linking errors. let me know so that I can compile a new package!
With sources it is different. You will find that often, the source for an older version of software will no longer compile on a newer Linux distribution. That is typically caused by updates to the gcc and glibc packages which introduce new library calls and interfaces which were not available at the time the older software was written. For successful compilation you would have to find patches or even use a newer version of the source. At the same time, a binary which was compiled from that old source, on an older Slackware release, will usually still work without issues on the new Slackware.
Eric
Hello Eric.
Just to report that qemu-kvm 1.2.0 does not work on 14.0 (missing libgnutls.so.26). I had to recompile it in order to make it to function.
Hi MartinOwaR
Thanks for mentioning this. I have just finished uploading some fresh qemu-kvm and vde packages, built on Slackware 14.
Cheers, Eric
Hi Eric,
Just a note to let you know about the release of IcedTea 2.3.3, see http://mail.openjdk.java.net/pipermail/distro-pkg-dev/2012-October/020571.html
And, once again, many thanks for all the work you do on Slackware!!
Cheers, Jean-Francois
Hi Jean-Francois
The new openjdk packages have been built already, and I will upload them soon. A blog article will have to wait until tonight.
Cheers, Eric
Hi Eric,
Thanks for the openjdk upgrade. I installed the new package and it works great!
Cheers, Jean-Francois
Just saying thanks for all of the work you have done. I use Slackware for my own network and for work related things. And the work you and the others do is appreciated.
thanx.
Thanks for the packages and updates. They work flawlessly ! Amazing stuff.
Hi Eric, the the *.info for your http://alien.slackbook.org/AST/ tool is out of date for SBo. The new .info file is updated to add a REQUIRES line, and removed APPROVED per http://slackbuilds.org/guidelines/
Hi JJ,
You are right, I fixed the .info generator. Thanks for notifying me.
Cheers, Eric
just wanted to say thank you for all your work.
Hello!
Want to write howto page in docs.slackware .com, registerd, log in, try to “add page”, and got
You are here: start » howtos » misc » sb_live_5.1
Permission Denied
Sorry, you don’t have enough rights to continue. Perhaps you forgot to login?
what we can do?
alienbob – Got one heck of a question for you. I’ve been tinkering, building a new system and in the name of keeping my wife happy, decided to take a stab at getting handbrake-gtk installed. So far, your build script is the only one I’ve found for the gui. Getting handbrake installed went without a hitch using a build script from elsewhere. But running your build script is causing my system to reboot.
What the heck? Reboot while building software? There’s something really strange going on. Any ideas? Is your build dependent on multi-lib or something that I’m not providing?
Any insight would be more than welcomed.
BTW, loved the calibre build. Nice work.
Hi OlPhart
My handbrake.SlackBuild does not need multilib or anything else. I run that script on a “virgin” Slackware in order to compile a package.
If your computer reboots during compilation then that might indicate an overheating issue. Software does not make your computer reboot just like that.
Is there anything in the message log right before the reboot occurs? Can you try to monitor the case- and CPU-core temperatures during the compilation?
Cheers, Eric
Thanks Eric ,
Temps are good. Don’t know what is going on but it’s certainly not temps that are doing it. Very abrupt reboot, no warning. Seems to occur at some point during ffmpeg build.
At the time I was getting the reboots I had been exec’ing the build script rather than sourcing.
When I try to source the script I get a rather cryptic kernel crash that freezes the system. Appears to be some sort of cpu related memory segmentation issue.
That said, the thought occured to me that a “virgin” build may be the answer. When I get the chance I’ll try a clean build and see how it goes.
When I get the chance to try a “clean” build I’ll get back to you and let you know how things go.
In the meantime, I’m off to shovel 6″ of fresh heavy wet snow.
Thanks,
Drew.
I brought this up in #slackbuilds a while back but thought I’d post it here because I wasn’t able to catch you online at the same time I was.
I wanted to suggest that –enable-libvorbis be added to your restricted ffmpeg slackbuild. While ffmpeg does use libvorbis internally, certain applications which rely on ffmpeg to transcode (Amarok’s transcoding dialog that it gives you when you put music on a portable player being the first one that comes to mind) appear to require –enable-libvorbis to have been passed during compilation in order to transcode to ogg.
I manually add it myself when I compile ffmpeg and it does not appear to cause any problems, and does not require any additional work for the user since libvorbis is included in a stock Slackware installation.
Anyway, just my suggestion. Thanks!
Cultist,
I will use “–enable-libvorbis” in future builds of my ffmpeg package. No problem at all.
Eric
Upgrading KDE to 4.10.x removed some important plasmoid SDK tools, not packaged into kde-workspace anymore (plasmaengineexplorer, plasmoidviewer etc.). Now they are included into Plasmate (http://dot.kde.org/2013/03/05/plasmate-10-plasma-workspaces-sdk). Could you try to build Plasmate slackware package?
Thank you for all your work
Tom
thank you.
Mr. slack_dragon
no requests. thank you for all!!
or maybe…..more tutorials to know the same as you know
Just one question: You have to replay mysql with mariadb in compat32 stuff, have you?
Best regards
Hi Michelino
Yes, I should have done that earlier.
Fixed and uploaded.
Cheers, Eric
Just found your modified inet1 files for bridging. Once again AB does the hard work so I don’t have to. Thank you, your a star.
Hi mickski
Those bridge modifications for rc.inet1 were merged into Slackware 14. Are you running an older version still?
Eric
Hey Bob
yeah still on 13.37, running kde 4.10.1 + lo 4 thanks to some slackbuilds I found somewhere :-). It really would be easier to just upgrade.
Couldn’t agree more about the awful weather.
Cheers 🙂
Hi Eric, I wonder if you have reached to compile wvdial and wvstreams on your arm port. Because now I’m having problems on my raspberry-pi trying tu use wvdial.
An error about:
getcontext open parenthesis &get_stack_return closes parenthesis == 0.
The error is not in the compile time, is in execution time. I didn’t have any problems compiling from the sources provide by slackbuilds.
Searching in google point out that this is and old problem,since 2009 and so on, but I have not read anything about some that fixed the problem.
Hi P431i7o
I have not yet compiled any slackbuilds.org package… just sticking to the official Slackware packages for now.
Perhaps this is your solution: http://lists.linuxtogo.org/pipermail/openembedded-devel/2011-February/030055.html
Eric
Eric
Hi, Is there any easy way to make VLC link to libva installed in the system?
What do I need to change in the slackbuild?
If you want to link to the system libva, then do not let the vlc.SlackBuild script enter the routine that builds an internal static copy of libva.
Eric
Hi Eric,
It appears there’s something weird going on with your Slackware mirror (taper). Whenever I try to check for updates in the patches directory with slackpkg, it reports no packages to be upgraded. However, switching to the slackbuilds.org mirror fixed the problem.
Also see: http://www.linuxquestions.org/questions/slackware-14/slackpkg-says-no-pkgs-to-update-after-installing-14-0-but-there-are-pkgs-in-patches-4175451800/
Hi JKWood
There was a small omission in my rsync_slackware_patches.sh script, some additional files were not being synced (FILELIST.TXT and CHECKSUMS.md5* in the main Slackware directory are updated whenever there are new patches).
I uploaded a fixed script and also ran that script on taper, so that the Slackware repositories there should work with slackpkg again.
Thanks for mentioning,
Eric
I just wanted to send you a small note to let you know how much I appreciate everything you do for the slackware community, and your continual efforts to make Slackware the most amazing distro available.
Your work is always impeccable, and I hope to someday be able to give back to the community as much as you have generously given over and over again.
Hi Bob.
I need help. I want to install Cairo Dock in slackware 12.0 and not locate binary packages. you know where I can find them?
Thank you.
Hi raymundo
I do not think you will find many binary 3rd party packages for Slackware 12.0 on the Internet. Most people only release for Slackware 14.0.
A SlackBuild for ciro-dock was added to http://slackbuilds.org/ for Slackware version 13.1. I have no idea if cairo-dock will work on an old Slackware like 12.0.
Eric
Thank you Bob.
Hi, Eric.
Appreciate all your hard work on Slackware and your repositories for that. Not sure if you use SSD disks. But for those who uses SSD disks and running encrypted LVM setup, there is a need to add option –allow-discards for the cryptsetup lusOpen commad in the init file for initrd.
Would appreciate if you update that.
Thank you in advance.
Hi dolphin77
Using “–allow-discards” has a potential negative security impact (see the cryptsetup man page). I guess that if you want to use this parameter because you installed Slackware on an encrypted SSD, you will have to add that parameter yourself to the init script inside the initrd. Slackware’s mkinitrd command won’t overwrite /boot/initrd-tree-init unless you specidy “-c” to (re)create the initrd from scratch.
Or you can submit a patch to introduce a variable which can be used to specify additional non-default cryptsetup parameters.
Eric
Eric, thank you for prompt reply.
You are right, I didn’t think of possible security impact. Probably it is better to leave as is system wide. Anyway on-line trimming (mounting with discard option) is not a good choice. Thus it is better to boot up from external flash disk from time to time and manually mount encrypted partitions and to run fstrim manually.
Thanks.
Hello, Eric
both 32-bit and 64-bit packages of kajongg-4.10.3 game are broken – there’s no executables in them.
greetz, Alek
Hi Alek
You are right. I checked the build logfile and it appears that Kajongg requires pyhon-twisted (a networking library) which again depends on zope-interface.
I do not think that zope will ever be integrated into Slackware, so I left a note for Pat to decide what to do about this.
Basically there are two options:
1) remove the kajongg package entirely from Slackware’s KDE package set
2) Force the installation of the kajongg binaries (after all they are only python scripts) and leave it to the Slackware user to install python-twisted and zope-interface from SlackBuilds.org (they are both present there already).
Eric
Thanks Eric, I’ll follow you forever!
Hi Eric,
I know that you’re concentrating on Samsung Chromebook. But, would your Arm port of Slackware run on BeagleBone Black ?
–William
Hi William
Unfortunately I can only test on hardware that I actually own. However, my ARM packages will work on any armv7 hardware. It’s usually just the kernel which you have to create for a new piece of hardware, and that will at least give you a bootable Slackware system. After that, there will usually be tweaks to get X.Org and sound fully functional but that will straighten out itself in time.
You could try what I initially did for my ChromeBook: find a bootable SD card image for another distro, write that to an empty SD card, and then wipe the other distro’s filesystem. Then copy my mini root filesystem into the empty partition and see if that will boot your BeagleBone…
A mini rootfs is here: http://taper.alienbase.nl/mirrors/alienarm/rootfs/armv7hl/chromebook/ . That rootfs does not contain a kernel or kernel modules, try copying those from a working BeagleBone distro image.
Eric
hi alienbob!
any chance to see an update of your handbrake package to v0.9.9?
alex
Hi, Eric.
Want to report that something is wrong with your slackware64-current mirror. Looks like new files were added during sync, the old ones were not deleted.
See for example http://taper.alienbase.nl/mirrors/slackware/slackware64-current/slackware64/kde/, where both kde-4.10.3 and kde-4.10.4 files are present.
Hi dolphin77
Yeah, I guess that Pat uploaded his KDE packages around the time that I run my mirror script. Usually those two events are far apart.
I re-ran my mirror script and all is well again.
Cheers, Eric
Hi Alexandre Jobin
Handbrake has been upgraded last week. It took some time to create a patch to make the compilation succeed.
Eric
Hi Erik, just installed calibre in slack 14.0 x86_64 multilib with all dependencies get this error.
$ calibre
Traceback (most recent call last):
File “/usr/bin/calibre”, line 20, in
sys.exit(main())
File “/usr/lib64/calibre/calibre/gui2/main.py”, line 415, in main
app, opts, args, actions = init_qt(args)
File “/usr/lib64/calibre/calibre/gui2/main.py”, line 85, in init_qt
from calibre.gui2.ui import Main
File “/usr/lib64/calibre/calibre/gui2/ui.py”, line 31, in
from calibre.gui2.widgets import ProgressIndicator
File “/usr/lib64/calibre/calibre/gui2/widgets.py”, line 21, in
from calibre.gui2.progress_indicator import ProgressIndicator as _ProgressIndicator
File “/usr/lib64/calibre/calibre/gui2/progress_indicator/__init__.py”, line 15, in
pi_error)
RuntimeError: Failed to load the Progress Indicator plugin: the sip module implements API v9.0 to v9.1 but the progress_indicator module requires API v8.1
Do I need to recompile in 14.0 or is something missing?
Thanx
Hi César
I thik you are not running Slackware 14. It seems that you added some newer software on top which replaced original Slackware packages. Are you running my KDE 4.10 for Slackware 14? Part of that is an upgrade to the Slackware ‘sip’ package which is incompatible with calibre’s Slackware 14 package.
You have to recompile calibre to fix the sip error.
Eric
Eric, yes I’m using your KDE 4.10.4. I’ll recompile calibre and try to make it work. Gracias
Hi Eric,
I’ve tried your alienarm miniroot fs on a device with NAND flash. I’ve created an ubifs image for rootfs and booted it. It always stops at the fsck check. ubifs does not support fsck at all, so I have patched rc.S to skip fsck for certain file systems: http://pastebin.com/AEJyy8qE
Thorsten
Hey Bob
Did you forget to update the libreoffice slackbuild on http://taper.alienbase.nl or do I just need some patience.
Thanks 🙂
mick, you looked at the “14.0” directories, right? I am not updating LibreOffice for Slackware 13.37, only for 14.0.
Eric
Hi Thenktor, interesting patch. In case your pastebin entry does not have eternal life, I copied its content to http://slackware.com/~alien/rc_scripts/other_rc_scripts/rc.S_selective_fsck.diff .
Was this a Nokia device you tried the mini rootfs on? What were your findings?
Cheers, Eric
Hey Bob
I mean the build script – http://taper.alienbase.nl/mirrors/people/alien/slackbuilds/libreoffice/build/libreoffice.SlackBuild it’s dated 18-may-2013.
last entry in the changes comments is :-4.0.3-1: 09/may/2013 by Eric Hameleers
# * New 4.0.3 release built for Slackware 14 and newer
The sources dir is up to date 30/6/13.
Once again many thanks 🙂
Lightening speed on the update Bob.
Thanks 🙂
ps
Don’t know where you find the time, but very much appreciated.
Hi Bob,
it was a Atmel devkit, based on a Cortex A5: http://www.atmel.com/devices/SAMA5D33.aspx
Except for the fsck thing I had no other problems. I’m using a kernel compiled from the Atmel source tree: https://github.com/linux4sam/linux-at91
Here is a picture: http://imageshack.us/photo/my-images/221/sy0l.jpg/?sa=0
I love slackware because I saw the birth of computer science. I’m sixty years and appreciates your work. It makes me very happy. I am Brazilian. Sorry my english. Congratulations!.
Frederico
Hey Eric! I couldn’t help but notice this – http://blog.documentfoundation.org/2013/07/18/the-document-foundation-announces-libreoffice-3-6-7/. Are you considering creating packages for Slackware 13.37, for old time’s sake? Either way, thank you for your continuous efforts to provide timely updates for all the extra software packages you offer to us fellow Slackers! 🙂
Jaycee, good idea. I will find some time for new LibreOffice 3.6.7 packages.
Eric
Hi Eric!.
I recently installed slackware64 (for the first time) on a Dell inspiron 14z, the thing is that it came with a hibryd video system, ati radeon plus intel graphics.
I proceeded to install it’s driver, but in some way that is described in many tutorials (download the driver, run sh amd-ati-installer.sh –buildpkg, that generated a tgz file). The thing is that it didn’t work, I tried with the beta release and the stable one, but not success, after that I looked forward into the Xorg logs, because the message was ‘no screen founds’ and something related to libgl not found.
So I started looking for that library and the result was that it was installed on /var/lib64 instead of /var/lib where it was looked for.
So the the thing is, who I have to report this? To Patrick or to the people of ATI?
I solved the problem just creating a link with ln -s of the folder fgl between the /var/lib and /var/lib64
Hi p431i7o
That sounds strange, the ATI driver looking for a file in /usr/lib on a 64-bit system which uses the lib64 convention for its library locations.
I would guess that this is an issue you should report to ATI.
Also, if you feel that this issue should be mentioned on http://docs.slackware.com/howtos:hardware:proprietary_graphics_drivers?s%5B%5D=ati#amd_ati_catalyst_driver_installation_and_configuration then you could use that article’s “Talk” page to document your findings and hope that a WIki editor picks it up.
Cheers, Eric
This isn’t a post to inform, complain, or anything like that.
It’s fairly simple.
Thanks, alienbob for all you do in the Slackware community. You are very appreciated by this guy —> Skinnable
Although the only Slackware installation currently running in my household is on a server, I would have been completely lost without all the posts provided by you on the many pages floating around the internet.
So, thanks so much for your help – and, I appreciate all you do.
Skinnable
Hey Eric! Another week, another LibreOffice release – http://blog.documentfoundation.org/2013/07/25/libreoffice-4-1-interoperability/. I was just wondering if you plan to create packages for Slackware 14.0, or wanted to wait for the release of Slackware 14.1. Considering that interoperability was a large focus of the release, it’d be handy for 14.0 users to have LO 4.1 packages available. Either way is fine though, I trust your judgement. 🙂
Hi Eric!
Here you can find a patch for building libtorrent-rasterbar 0.16.10 with boost 1.54 (as in current): https://projects.archlinux.org/svntogit/packages.git/tree/trunk?h=packages/libtorrent-rasterbar
With respect to your desire to have a tablet with Slackware on it, have you heard of the PengPod. I would love to have tablet with Linux on it but as I am not up to date on the politics of computer freedom, I find myself bouncing back and forth between the pengpod site and some of your articles on a port to the Arm architecture. Just Curious what your thoughts may be on the topic. Thanks.
I’ve followed your work for quite awhile, and as a last resort as not to waste your time I’ve come to you. Somewhere down the line slackpkg has been corrupt on a newly installed Slack 14 system. I’ve reinstalled the system as well as slackpkg to no avail, I’m more then willing to open a SSH tunnel for you if you would be willing to assist in either assisting me to resolve this mattery or I’m more then willing to follow any and all direction.
Thanks
Hi Austin
This blog is no replacement for the user forum at linuxquestions.org … problems like the one you describe are hard to diagnose without more than “slackpkg has been corrupt”.
What I suggest you do, is open a thread on http://www.linuxquestions.org/questions/slackware-14/ and describe your problem there. Please try to give more information than what you wrote here, because I can not even start diagnosing it. Error messages? Weird behaviour? How do you experience this corruptness? Did stuff stop working? Etc…
I do not perform one-on-one support, so the ssh tunnel offer is nice but I am not going to follow up on it. I hope you understand that.
Cheers, Eric
Hi Charles
I looked at the PengPod but I think its hardware is a generation too old for the price tag. Nice initiative though and I wish there were more like this. Tablet computer is not going away but the closed hardware / closed driver approach of most vendors _really_ annoys me.
Cheers, Eric
Hi,
Eric, condolences for the lose of your dear friend ‘Sox’. I too lost a friend a few weeks back who would be at my side when working on the farm. ‘Mo’ will be remembered and loved for her unconditional love. Best dog!
-Gary
Thanks for your response. Yes, I just wish there was a tablet that had a basic linux os like slackware so I could practice writing scripts or programs. In some ways this closed hardware fight reminds of the early 90’s but from a different standpoint. The drivers didn’t exist, at least not the interconnectivity of machines. I still have my Macintosh Quadra 605 with 4 mb of ram 80 Mb HD, that I programmed in assembly, and pascal and c but that was a a long while back. I used a modem to connect to the university and did my work on a Unix mainframe. Those were the days. Appreciate greatly your response. Cheers.
Hi Eric.
Would it be possible to add support for OpenConnect VPN to the networkmanagement package in KDE?
I successfully built the latest openconnect (5.01) and NetworkManager-openconnect (0.9.8.4) using the slackbuilds from slackbuilds.org (just had to bump version numbers). Unfortunately OpenConnect did not appear in the VPN list in the NetworkManager configuration. It seems the networkmanagent package depends on openconnect to be installed to build support for it. After installing openconnect I manually built the networkmanagent package using the KDE.SlackBuild and sources from the slackware64-current repo (4.10.5). Now it’s working like a charm. I guess this would add OpenConnect as a dependency to the KDE build.
I need openconnect because my college uses Cisco AnyConnect VPN and it failed to install via the browser. Earlier I got OpenVPN working after installing NetworkManager-openvpn using Robby Workman’s slackbuild so I was led to believe the same would apply to NetworkManager-openconnect. Googling around only confused me but in the end I found the answer in the sources (networkmanagent-0.9.0.8/vpnplugins/openconnect/CMakeLists.txt).
Cheers, Marius.
Hi Marius
I think the best course of action for you would be to write to Pat Volkerding (volkerdi at slackware dot com) asking for the inclusion of openconnect and NetworkManager-openconnect into Slackware.
I am carefully aiming my KDE stuff at slackware-current and therefore not introducing new packages unless absolutely required for KDE. Openconnect is a bit of a grey area.
The build-time dependency of networkmanagement on the presence of openconnect is unfortunate.
Eric
Thanks for your reply.
I followed your advice and asked Pat Volkerdi to include openconnect into Slackware. I also wrote to Robby Workman because he is the maintainer of the openconnect and NetworkManager-openconnect packages at slackbuilds.org and there is no mention of this issue there.
Cheers, Marius.
Hi Eric,
I was wondering what happened to the URL
http://www.slackware.com/~alien/samba4/
I get an error 404. Did it move or get abandoned?
– dr –
Hi Eric,
I’m a recent Slackware convert. Thanks for setting up this blog and putting up so much valuable information for sharing.
Recently, I wanted to learn how to compile my own kernel and I followed your post here:
http://alien.slackbook.org/dokuwiki/doku.php?id=linux:kernelbuilding&s%5B%5D=kernel&s%5B%5D=build
unfortunately, I made a very silly mistake when creating the linux symlink. Instead of running:
#ln -s /usr/src/ linux
I did this:
#ln -s /usr/src// linux
with an additional “/”
hence, my symlink linux -> /usr/src///
notice the double “//” at the end.
Upon boot up, I get a whole series of version magic error:
version magic ‘3.2.29-custom SMP mod_unload PENTIUMIII ‘ should be ‘3.2.29-custom SMP mod_unload 486 ‘
upon realising my mistake, I rm the symlink and source dir and start everything from scratch. However, the error persisted even though I select and rebuild all my modules with Pentium III settings.
Could I have corrupted my entire system with the wrong linux symlink?
Is there a big difference between
linux -> /usr/src// and
linux -> /usr/src///?
Appreciate if you can provide some insights to this puzzle.
Thanks in advance for your generousity!
ck
Hi Eric,
Sorry that I had to repost this as I realised all my comments within “” were somehow truncated.
I’m a recent Slackware convert. Thanks for setting up this blog and putting up so much valuable information for sharing.
Recently, I wanted to learn how to compile my own kernel and I followed your post here:
http://alien.slackbook.org/dokuwiki/doku.php?id=linux:kernelbuilding&s=kernel&s=build
unfortunately, I made a very silly mistake when creating the linux symlink. Instead of running:
#ln -s /usr/src/linux-3.2.29-custom linux
I did this:
#ln -s /usr/src/linux-3.2.29-custom/ linux
with an additional “/”
hence, my symlink linux -> /usr/src/linux-3.2.29-custom//
notice the double “//” at the end.
Upon boot up, I get a whole series of version magic error:
version magic ’3.2.29-custom SMP mod_unload PENTIUMIII ‘ should be ’3.2.29-custom SMP mod_unload 486 ‘
upon realising my mistake, I rm the symlink and source dir and start everything from scratch. However, the error persisted even though I select and rebuild all my modules with Pentium III settings.
Could I have corrupted my entire system with the wrong linux symlink?
Is there a big difference between
linux -> /usr/src/linux-3.2.29-custom/ and
linux -> /usr/src/linux-3.2.29-custom//?
Appreciate if you can provide some insights to this puzzle.
Thanks in advance for your generousity!
ck
Hi donrich39 –
Samba4 was eventually added to Slackware-current and therefor my own samba4 packages were no longer needed. But I keep a copy at http://taper.alienbase.nl/mirrors/people/alien/samba4/
Eric
Hi ck
Apparently, the kernel you boot and the modules you are trying to load are not built from the same sources.
The error you posted tells you that the modules which are being loaded are compiled for “PENTIUMIII” CPU architecture, while the kernel has been compiled for “486”
The extra slash in the symlink should not have any effect on this.
You need to recompile your kernel plus modules and install them all. Perhaps you forgot to install the new kernel which you built for PENTIUMIII ?
Eric
Hi Eric,
Thanks for your advice.
I double checked to confirm that the kernel was build for pentium III under the processor config. Still same error.
So I rm and cp new source for the build using config seed from kernel-seeds.org.
It works now. Thanks again!
cheers
ck
hi eric,
just had a question since i use your build of ffmpeg, is that includes the libavcodec also? ( I’m not sure but i would say it was in your ffmpeg, since after upgrade to 12-Jun-2013, i get the error that it is missing.)
thank you,
cheers,
inman.
Hi inman, what program is giving you issues? It’s most likely not FFmpeg but a progranm which uses FFmpeg.
Check the exact error. It will probably say something like “libavcodec.so.53: not found”.
My latest FFmpeg package contains “libavcodec.so.54”. Any program which you have compiled against an older version of FFmpeg will stop working because the SONAME of that library has changed. FFmpeg does that a lot.
This is the reason why I bundle an internal copy of FFmpeg with the VLC package for instance… to avoid this kind of breakage.
Eric
Hi Eric,
Thanks for referring me to slack-current and the samba 4 pkg. Upgraded to current, installed samba 4, needed kerberos5 so got the slackbuild from slackbuilds.org and the build failed. The reason is the tcl package has been upgraded to v.8.6 in slack-current
and some test code in the kerberos5 package uses (Tcl_Interp *) interp->result, which is deprecated in Tcl 8.6
and only defined when USE_INTERP_RESULT is defined.
The fix is to breakout the offending file (tcl_kadmin5.c) and add the line:
#define USE_INTERP_RESULT
before the #include declaration. I.E.:
/* -*- mode: c; c-basic-offset: 4; indent-tabs-mode: nil -*- */
#include “autoconf.h”
#include
#include
#if HAVE_TCL_H
#define USE_INTERP_RESULT
#include
#elif HAVE_TCL_TCL_H
#include
#endif
#define USE_KADM5_API_VERSION 2
#include
#include
#include
#include
#include
#include “tcl_kadm5.h”
then replace the file in the archives and rerun the slackbuild.
I can give the commands if anyone else has the problem and is interested.
Thanks, – dr –
For some reason in my above post, all the include statements had the greater-than “filename” less-than info removed.
the #define statement had to be before the #include tcl.h statement.
Mr. Hameleers, I wanted to thank you for all the documentation and support you provide for Slackware. I am writing today to see if I might be of some assistance. I discovered something neat while trying to install Slack on Apple products that might help others who encounter similar issues as myself.
A while back, I was offered a chance to purchase an Apple Mac Pro for a crazy-reasonable price. I have never been a fan of over-priced, last-years hardware; however, this machine had really great specs and a very attractive price. I have quite a few systems in my home, for various purposes; however, I have always wanted a true dual CPU machine. So, I went ahead and bought the Mac. As expected, I wasn’t too enthralled with Mac OS. It had some really nice features; but overall, it seemed lacking in some features that now seem basic. I began to experiment with the Mac and noticed that I couldn’t boot most install disks on it. After working with, and learning, the ups and downs of the “Mac” EFI implementation, I discovered something.
Out of all the distro’s I tried, the Ubuntu disk was the only one that started up and ran normally. It ran well, even in live mode. I did a little digging and started looking at the version of Grub on the Ubuntu boot disk, and it was unique. I tried many versions of Linux looking for one that gave me choice and also had a simple interface at boot. I found that Slackware was the best choice. Unfortunately, I couldn’t get the system to boot with the Slackware64-14 disk. I believe I read somewhere that the Linux Kernel 3.2.29 in that version wasn’t yet compatible with EFI. So I cloned Slack64 current from a mirror, built an ISO and tried to boot. No joy. So then I started playing around with the images. After much trial and error, I was able to extract both images to a folder and make a working installer. I found that the Grub.cfg for the Ubuntu version loaded additional add-ons that the Slackware version did not. Some are realted specifically to EFI and Apple products. So I took the Slack, Grub.cfg and then the Ubuntu, Grub.cfg and made a hybrid. The only difference is that the new one includes the additional add-ons that the Ubuntu installer used. I replaced the version of grub (from the Slack disk) with the one included with Ubuntu and it worked. I was able to install from DVD as well as USB. I have pretty much perfected this method. I know that many would ask why would someone want to run Slackware on a Mac. Well, because it’s awesome, and because you can. Also, older Macs get no love or support from Apple. Ultimately, my main goal was to compare features and the usage of system resources. I also wanted to test what compiling with Dual Quad-Core Xeons with 16GB DDR3 ECC would be like. I have to say, very nice. Both CPU’s support hyper-threading. If I specify 16 threads, I can compile a kernel in about 3 minutes. Anyways, sorry for the long email. I just wanted to get this information to you. If you think this method could be helpful, please let me know and I will make a tut showing what I did so others can do the same. Also, I noticed this method worked with my other UEFI systems as well. Might help some other folks having issues.
I was looking for an experience with Linux that I was unable to find other places. I want to learn the details of the OS and what makes it tick. Slackware forced me to learn in order to get what I wanted, and ultimately, that is exactly I needed.
Thanks,
Austin
Hi Austin
I do not think that this information you are able to supply will find its way into Slackware 14.1 (the doors are pretty much closed on that) but why don’t you request an account at http://docs.slackware.com/ and write a nice tutorial about what you had to do to Slackware’s grub in order to boot a Mac Pro?
The Slackware Documentation Project would be the ideal location for hosting such information.
Cheers, Eric
I kinly like to ask for a Slackware Mini 14.1 build. Thank you!
Hi Mr.X
Uploading mini ISO images now.
Eric
I believe there is a general problem with vlc 2.1 tearing with x264 videos. Do you still have a 2.0 build somewhere until they fix it?.
Hi Boris,
I keep a copy of old VLC 2.0.8 packages here: http://slackware.com/~alien/vlc20/
Eric
Hello. I would like to install latest transmission 2.82 n my current 64 box. Somehow it fails to compile when it comes to qt- part of compiling. If i comment qt-part on sb.script it goes well, but I would like to have qt as well. It fits nice to kde. Thanks.
Hi Chu_Set
I do not like Transmission myself, I use qbittorrent which is also Qt-based and is a much cleaner program: http://slackware.com/~alien/slackbuilds/qbittorrent/ .
But if you want to compile transmission, give this SlackBuild script a try, it is checked on Slackware 141:
https://github.com/Ponce/slackbuilds/blob/current/network/transmission/transmission.SlackBuild
Eric
Are you going to rename current/ to 14.1/ at http://alien.slackbook.org/ktown/?
hey, how much do I have to click on the ads? cheers..
I saw that you packed a screen videocapture program called xvidcap a while ago. Is there any reason why xvidcap-1.1.7-i486-1alien.tgz would not work on a 14.0 installation: http://www.slackware.com/~alien/slackbuilds/xvidcap/pkg/12.1/
Sorry, I really love that package. I used to work with it before. It just made the job. I work on porteus 2.1.
Thanks.
Hi dmitri,
Next KDE is going to go into a “14.1” directory instead of current.
Eric
Hi Ferdi
Don’t stress yourself 🙂
Just click from time to time. Google’s adwords program recognizes abnormal click behaviour and discards those click-through actions…
Eric
Hi francois.e,
Perhaps I should just build a new xvidcap package for Slackware 14.0 / 14.1.
Eric
Just FYI, dconf-0.18.0-x86_64-1.txz, is included in -current.
🙂
Hi cwizardone
Yeah you’re right. I have removed the redundant package from the repository.
I also patched gecko-mediaplayer so that it works with the Chromium & Chrome browsers
Eric
Eric,
I have used your multilib packages for some time. THANK YOU! However I recently installed MesaLib-9.2.3 on slackware 14.1 (graphics issues) and I am now ‘in between’ the 64 bit mesa library for 9.2.3 and the 32 bit library for the compat32-9.1.7 library. Is there a way that I can get both these libraries in sync?
Thanks, jwc
Hi john,
What you can do is compile a 32-bit package for that MesaLib-9.2.3 (should be doable on your multilib computer or else ask someone with a 32-bit Slackware installation) and then use ‘convertpkg-compat32’ on the 32-bit package.
Eric
Eric, What command do I use to get a 32 bit compile? The plain vanilla configure and make commands give the the 64 bit version. Incidentally the resulting installation has gone into /usr/local/lib instead of /usr/lib. I don’t know if that is a problem or not, but is there a simple way to change it?
Thanks, John
Hi John
The best way forward would be to remove the Mesa stuff in /usr/local (you can probably just run “make uninstall” in the source directory) afrter creating a proper new 64-bit mesa package.
How to do that:
* grab the mesa source directory from Slackware-current, edit the mesa.SlackBuild and change the VERSION number, and place the new MesaLib source tarball inside that same directory.
* Then, run ./mesa.SlackBuild in that directory which will compile a mesa-9.2.3 package that will be placed in /tmp/
* Use the command “upgradepkg –install-new /tmp/mesa-9.2.3-*.t?z” to install your new mesa package (upgrading anything that remains from the original Slackware mesa package)
* Then proceed building a 32-bit package for mesa. Follow the guidelines here to accomplish that: http://docs.slackware.com/slackware:multilib#compiling_32-bit_programs
* Finally, convert the 32-bit package to a “compat32” package and upgrade your computer with the result. Something like:
# convertpkg-compat32 -i /tmp//mesa-9.2.3-i486-1.txz
# upgradepkg –install-new /tmp//mesa-compat32-9.2.3-x86_64-1compat32.txz
And do not forget to re-install your binary (Nvidia or Ati) graphics driver every time you upgrade the mesa package!!
Eric
“grab the mesa source directory from Slackware-current, edit the mesa.SlackBuild and change the VERSION number, and place the new MesaLib source tarball inside that same directory.
* Then, run ./mesa.SlackBuild in that directory which will compile a mesa-9.2.3 package that will be placed in /tmp/”
The mesa package that I get when I do that has all the lib files in /usr/local/lib and the header files in /usr/local/include. I don’t know how to change the slackBuild file so the files go to the right place.
Do you know if anyone has written a more recent slackbuild script for rng-tools? The only one I find is from Slackware 12
Hi q5sys
This is a question about the rng-tools entry on slackbuilds.org , right? You should really be asking that on the slackbuilds-users mailing list or visit the #slackbuilds IRC channel on Freenode to get an answer.
Nobody seemed to care about maintaining the SlackBuild script, why don’t you volunteer?
Eric
Hi Eric,
I was introduced to Slackware and Linux on general when I built my unRAID server, and I firstly want to say a huge thank you for all your packages and scripts I’ve used the last couple of years!!
I am a complete noob however when it comes to writing a slackbuild… I usually rely on botching together someone’s old script and thats what I ahve done for Makemkv(con).. until now! Theres a new dependency (libavcodec) required from ffmpeg, and I’m at a complete loss how to get that to compile for Slackware 13.1 (unRAID) to get the latest MakeMKV to compile with my (botched) slackbuild script. Any pointers would be massively appreciated!!
Alex
Alex, libavcodec is part of ffmpeg. Have you tried compiling my ffmpeg.SlackBuild on your Slackware 13.1 system?
Btw – any reason why you are sticking with a relatively old Slackware version? Chances get bigger with every release that older releases no longer support compilation of newer software.
When the ffmpeg package is installed, other software should be able to pick up libavcodec.
Eric
Thanks for your reply Eric!
unRAID is built on Slackware 13.1 ( or possibly 13.37) so unfortunately I’m stuck with it!
I tried your ffmpeg.Slackbuild tonight and it failed unfortunately :/. My laptop is busy running chkdsk so can’t tell you exactly what it failed on other than I remember it failed in line 1557!
I’ll check what the actual failure was in the morning! If you ever fancy making a MakeMKV package in the meantime, I’d never complain! Haha. In all honesty tho I wouldn’t mind getting my head around it all a bit better!
Alex
Eric,
Is there some reason why http://taper.alienbase.nl/mirrors/slackware/slackware64-current/ is not being updated? I’ve had my mirror-slackware-current.sh pointed there for a while, but since the upgrade, it hasn’t shown any movement.
Regards & thanks for all the effort,
Bill
Oops!
Looks like after the crash on 5 november, a lock file was not removed and therefore the mirror process was stuck. I have deleted the lockfile and manually started the mirror script. It should be OK in a short while.
Thanks for reporting this.
Eric
Hi Eric,
Sorry for the late reply! Still having issues and any help would be greatly appreciated!
I’ve installed your latest FFmpeg pkg ( 13.37 version) and I’m running the slackbuild script from here:http://slackbuilds.org/repository/13.37/multimedia/makemkv/ (Version Number changed).
I keep getting the error “checking LIBAVCODEC_VERSION_MAJOR… failed
configure: error: in /tmp/SBo/makemkv-oss-1.8.7′:
configure: error: LIBAVCODEC_VERSION_MAJOR is not known at compile time in libavcodec.h
See config.log’ for more details”
There is a libavcodec.pc file in /usr/lib/pkgconfig so unsure what the problem is?! For info the build instructions for MakeMKV is here: http://www.makemkv.com/forum2/viewtopic.php?f=3&t=224
“convertpkg-compat32” converts incorrectly “lesstif-0.95.2-i486-1.txz” package. “installpkg” throws message:
“install/doinst.sh: line 23: syntax error: unexpected end of file”.
Keyword “fi” is missing (line 5) in doinst.sh script…
Hi Rysio
Yes I am aware of that, it has always been there since the first multilib packages.
But I did not care to write a shell script parser just for the converting… and the error is harmless.
Eric
Hi Eric. It has been a long time since I visit your site. I feel I should spend more time here and with you. How do you join the blog? I can’t find a register button anywhere. I’m getting old (62 tomorrow).
Thank you.
Hi moonstroller.
There is no need to register to this site, posting comments is always allowed.
If you want to be kept informed of new posts you can use a RSS feed reader (thunderbird will do the job) and subscribe to the RSS feeds for my posts and/or comments (the links are at the bottom of every page).
Happy birthday!
Cheers, Eric
Hello, Eric!
First off, THANK YOU for all of your hard work. I’ve been using Slackware Linux for over 12 years now, and what you’ve done for Slack makes the overall Slackware experience that much quicker, easier, and enjoyable.
THANK YOU.
So, I’m writing to let you know that a problem seems to exist in your new LibreOffice 4.1 packages when attempting to start/play a slide show in Impress (Pressing F5 to start the show to present the slides in full screen mode). The screen freezes and will not present the slides. Pressing ESC will get you out of the freeze, but you still cannot present.
Happy Holidays.
lbs, that is actually not a new problem.
See here, google will show more hits: http://www.linuxquestions.org/questions/slackware-14/impress-libreoffice-4-1-2-f5-freeze-on-slackware-current-4175480364/#post5043725
Basically:
– Go to “Tools > Options > LibreOffice (View)”
– Untick “Use Hardware Acceleration”
– Restart LibreOffice Impress
Eric
Hi Alien Bob I just wanted to take some time out to say a big thank you to you I have used slackware now since about 1998/9 and built lots of my own pkgs over the years, then children arrived and my time diminished your packages and fixes have saved me so many hours of work time I have used well. I just found your cure for dropbox now I can make real use of it from my linux systems. I run your firewall to jolly nice and secure BT my isp said we cant see any thing you have connected to our system thats good I said Alien Bobs firewall’s working a treat gob smacked they were lol. any way have a great 2014 when the kids are a bit bigger I will be back in the circles time I gave back to slackware. Sam
Hi Sam
Good to see a Slacker who is into audio equipment servicing & repairs. Perhaps the Studioware guys should check you out (and vice versa)…
Cheers, Eric
How about infinality fonts patch for Slackware 14.1?
I am not interested in those Infinality font patches since they are mostly targeting MS WIndows font usage. If you use the default Slackware open source fonts you should not have a need for these patches IMHO.
Eric
Actually you can target Linux too or even Mac OS X fonts (bash /etc/fonts/infinality/infctl.sh setstyle). Anyway keep up the good work.
Hi,
Are you using Slackware in virtualised environment (most precisely in Citrix XenServer)?
I’m trying to install XenServer Tools in Slackware 14.1, but unfortunately Slackware is not officially supported.
If you have any experiences with this, or if you know anybody running Slackware on XenServer, please let me know. (I already tried to ask for help on linuxquestions.org, but I got no response.)
Thanks,
kukukk
Hi kukukk
I only use Slackware in VirtualBox and in QEMU. I have never tried Citrix Xenserver.
I found your post: http://www.linuxquestions.org/questions/slackware-14/installing-xenserver-tools-on-slackware-14-1-a-4175488000/ but I thik there was no response because of the lack of detail. Without a copy/paste of the commands you used and the responses/errors you got there is nothing sensible to answer.
Eric
Hi, and thanks for the quick reply.
I was almost sure that when you are creating/testing packages for different Slackware versions, you use them in a virtualised environment, and I hoped that it’s XenServer.
I did not added any other details because I don’t have :}. Converting and installing the Red Hat package went without any error message, but after restarting the guest XenServer still complained about missing XenServer Tools. Probably it’s because the different internal structures of the systems (something is not started, is not the right folder, etc). I don’t know, unfortunately I’m not a Linux expert. I hoped that somebody already went trough this and knows the steps required for installing XenServer Tools.
Anyway, Slackware is working fine as guest in Citrix XenServer, I just don’t have some information in the Management Console, like memory and cpu usage.
Alternatively, try installing that RPM directly in Slackware, not converting it to a Slackware package first. That way, the RPM pre- and post-install scripts will be executed.
Use the command “rpm –nodeps” to avoid getting a ton of missing dependency errors.
Eric
Hi Eric,
I hope you can help me to find out what is wrong with my installation of pipelight.
I have followed the instructions from http://alien.slackbook.org/blog/pipelight-using-silverlight-in-linux-browsers/.
when I run Pipelight diagnostic page, I get:
Pipelight diagnostic:
Please select the Plugin you want to test:
User agent (Javascript)
Checking for Windows user agent …okay
Mozilla/5.0 (Windows NT 6.1; WOW64; rv:24.0) Gecko/20100101 Firefox/24.0
Please note: not every user agent works on every site, try multiple ones if something doesn’t work!
Silverlight (as seen by a website)
Checking for Silverlight …failed
Pipelight
Checking for Pipelight …okay
Configuration of Pipelight
Checking if config exists …okay
/usr/share/pipelight/pipelight-silverlight5.1
Checking if pluginLoaderPath is set and exists …okay
/usr/share/pipelight/pluginloader.exe
Checking if winePath is set and exists …okay
/usr/libexec/wine-pipelight/bin/wine
Checking if wine exists …okay
/usr/libexec/wine-pipelight/bin/wine
Checking if winePrefix is set and exists …okay
/home/duo/.wine-pipelight/
Checking if dllPath/dllname is set and exists …failed
You need to define either a valid dllPath/dllName or regKey
Unable to verify if the DLL exists, please check this manually!
(dllPath = c:\Program Files\Silverlight\5.1.20913.0\)
(dllName = npctrl.dll)
Checking if dependencyInstaller is set and exists …okay
/usr/share/pipelight/install-dependency
Checking if dependencies are defined …okay
Distribution
Loading file /etc/issue …okay
Welcome to \s \r (\l)
Content of file: /usr/share/pipelight/pipelight-silverlight5.1
Loading file /usr/share/pipelight/pipelight-silverlight5.1 …okay
#
# Enables a diagnostic mode which might be helpful to find an
# error in the configuration or installation.
# To get the error messages go to: http://fds-team.de/pipelight
#
# Information for advanced users: The diagnostic page embeds the plugin
# of type “application/x-pipelight-error” to trigger the output of some
# debug information. The plugin cannot be triggered if everything is
# working, so this only affects users with a broken installation.
# The debug output will include paths on the local filesystem and the
# linux distribution used. If you don’t want to leak this information
# accidentially in case of a broken installation please either uninstall
# Pipelight or disable the diagnosticMode. [default: false]
#
diagnosticMode = true
#
# Path to the wine directory or the wine executable. When you
# specify a directory it should contain /bin/wine.
#
winePath = /usr/libexec/wine-pipelight
#
# Path to the wine prefix containing Silverlight
#
winePrefix = $HOME/.wine-pipelight/
#
# The wine architecture for the wine prefix containing Silverlight
#
wineArch = win32
#
# DLLs to overwrite in Wine
# (prevents Wine from asking for Gecko, Mono or winegstreamer)
#
wineDLLOverrides = mscoree,mshtml,winegstreamer,winemenubuilder.exe=
#
# Path to the plugin loader executable
# (Should be set correctly by the make script)
#
pluginLoaderPath = /usr/share/pipelight/pluginloader.exe
#
# Path to the runtime DLLs (libgcc_s_sjlj-1.dll, libspp-0.dll,
# libstdc++-6.dll). Only necessary when these DLLs are not in the same
# directory as the pluginloader executable.
#
gccRuntimeDlls =
#
# Path and name to the Silverlight directory
# You should prefer using regKey to make it easier to switch between
# different versions.
#
dllPath = c:\Program Files\Silverlight\5.1.20913.0\
dllName = npctrl.dll
#
# Name of the registry key at HKCU\Software\MozillaPlugins\ or
# HKLM\Software\MozillaPlugins\ where to search for the plugin path.
#
# You should use this option instead of dllPath/dllName in most cases
# since you do not need to alter dllPath on a program update.
#
# regKey = @Microsoft.com/NpCtrl,version=1.0
#
# fakeVersion allows to fake the version string of Silverlight
# Allows to get around some version checks done by some websites
# when using an old version of Silverlight.
#
# fakeVersion = 5.1.20913.0
#
# overwriteArg allows to overwrite/add initialization arguments
# passed by websites to Silverlight applications. You can
# use this option as often as you want to overwrite multiple
# parameters. The GPU acceleration state of Silverlight can be controlled
# by setting:
#
# enableGPUAcceleration=false # disable GPU acceleration
# comment out # let the application decide (default)
# enableGPUAcceleration=true # force GPU acceleration
#
# You may need to overwrite the minimum runtime version if
# you use an old Silverlight version as some websites set
# an artificial limit for the version number although it
# would work with older versions.
#
# overwriteArg = minRuntimeVersion=5.0.61118.0
# overwriteArg = enableGPUAcceleration=false
# overwriteArg = enableGPUAcceleration=true
#
#
# windowlessmode refers to a term of the Netscape Plugin API and
# defines a different mode of drawing and handling events.
# On some desktop enviroments you may have problems using the
# keyboard in windowless mode, on the other hand the drawing is
# more efficient when this mode is enabled. Just choose what works
# best for you. [default: false]
#
windowlessMode = false
#
# embed defines whether the Silverlight plugin should be shown
# inside the browser (true) or an external window (false).
# [default: true]
#
embed = true
#
# Path to the dependency installer script provided by the compholio
# package. (optional)
#
dependencyInstaller = /usr/share/pipelight/install-dependency
#
# Dependencies which should be installed for this plugin via the
# dependencyInstaller, can be used multiple times. (optional)
#
# Useful values for Silverlight are:
#
# -> Silverlight versions (you need to adjust dllPath):
# wine-silverlight5.1-installer
# wine-silverlight5.0-installer
# wine-silverlight4-installer
#
# -> optional depependencies (required by some streaming sites)
# wine-mpg2splt-installer
#
dependency = wine-silverlight5.1-installer
dependency = wine-mpg2splt-installer
dependency = wine-wininet-installer
#
# Doesn’t show any dialogs which require manual confirmation during
# the installation process, like EULA or DRM dialogs.
# [default: true]
#
quietInstallation = true
#
# In order to support browsers without NPAPI timer support
# (like Midori) we’ve implemented a fallback to
# NPN_PluginThreadAsyncCall. In the default configuration
# a timer based approach is preferred over async calls and the
# plugin decides by itself which method to use depending on the
# browser capabilities. Setting the following option to true
# forces the plugin to use async calls. This might be mainly
# useful for testing the difference between both event handling
# approaches. [default: false]
#
# eventAsyncCall = true
#
# The opera browser claims to provide timer functions, but they
# don’t seem to work properly. When the opera detection is
# enabled Pipelight will switch to eventAsyncCall automatically
# based on the user agent string. [default: true]
#
operaDetection = true
#
# Minimal JavaScript user agent switcher. If your page doesn’t check
# the user agent before loading a Silverlight instance, you can use
# this trick to overwrite the useragent or execute any other Java-
# Script you want. You can use this command multiple times.
# Uncomment the following 4 lines for FF15 spoofing.
#
# executejavascript = var __originalNavigator = navigator;
# executejavascript = navigator = new Object();
# executejavascript = navigator.__proto__ = __originalNavigator;
# executejavascript = navigator.__defineGetter__(‘userAgent’, function () { return ‘Mozilla/5.0 (Windows NT 6.1; WOW64; rv:15.0) Gecko/20120427 Firefox/15.0a1’; });
#
# We are currently implementing hardware acceleration support, which
# can cause problems on some systems as especially specific open source
# drivers render only garbage when hardware acceleration is enabled.
# To prevent breaking any working solutions, we are currently
# implementing a whitelist system, which will enable hardware
# acceleration by default if the specified shell scripts returns 0.
# Otherwise we will disable it. You can still use
#
# overwriteArg = enableGPUAcceleration=false/true
#
# to overwrite the check results. If you really want to skip this test
# you can use: silverlightGraphicDriverCheck = /bin/true
#
silverlightGraphicDriverCheck = /usr/share/pipelight/hw-accel-default
#————————- EXPERIMENTAL ————————-
# Watch out: The following section contains highly experimental
# stuff! These functions are likely not working properly yet and
# might be removed at any time.
#
# Silverlight uses a lot of timer stuff do to the window redrawing
# . In order to speed this up a bit the following switch enables
# some API hooks to do most of timer stuff in user mode (without
# having to call wine-server each time). It is still unclear
# if this option has any significant effect on the performance.
# [default: false]
#
# experimental-userModeTimer = true
#
# In order to make it possible to let a window stay opened in fullscreen, even
# if the user clicks somewhere else it is necessary to install a window class
# hook. With some plugins this could lead to other problems! [default: false]
experimental-windowClassHook = true
#
# A sandbox is a method to isolate an untrusted program from the rest of
# the system to prevent damage in case of a virus, program errors or
# similar issues. We’ve been developing the ability to use a (self-created)
# sandbox, but this feature still has to be considered experimental.
# The feature will only be used when the sandbox path exists.
#
sandboxPath = /usr/share/pipelight/sandbox
On the terminal window running firefox I get:
[PIPELIGHT:LIN:unknown] attached to process.
[PIPELIGHT:LIN:unknown] checking environment variable PIPELIGHT_SILVERLIGHT5_1_CONFIG.
[PIPELIGHT:LIN:unknown] searching for config file pipelight-silverlight5.1.
[PIPELIGHT:LIN:unknown] trying to load config file from ‘/home/duo/.config/pipelight-silverlight5.1’.
[PIPELIGHT:LIN:unknown] trying to load config file from ‘/etc/pipelight-silverlight5.1’.
[PIPELIGHT:LIN:unknown] trying to load config file from ‘/usr/share/pipelight/pipelight-silverlight5.1’.
[PIPELIGHT:LIN:silverlight5.1] GPU driver check – Your driver is supported, hardware acceleration enabled.
[PIPELIGHT:LIN:silverlight5.1] using wine prefix directory /home/duo/.wine-pipelight/.
[PIPELIGHT:LIN:silverlight5.1] checking plugin installation – this might take some time.
[PIPELIGHT:LIN:silverlight5.1] basicplugin.c:373:checkPluginInstallation(): error in execvp command – probably dependencyInstaller/sandbox not found or missing execute permission.
[PIPELIGHT:LIN:silverlight5.1] basicplugin.c:383:checkPluginInstallation(): Plugin installer did not run correctly (exitcode = 1).
[PIPELIGHT:LIN:silverlight5.1] basicplugin.c:142:attach(): plugin not correctly installed – aborting
Same problem happens when I change silverlight plugin version.
Cheers, Duodecimo.
Hi Eric,
Oops, I forgot to mention on the previous message that I use Slackware 14.0 64 bits multilib.
Cheers, Duodecimo.
Hi Duodecimo
Did you execute the command (under your own user account):
pipelight-plugin –enable silverlight
Also, you should check the enabled plugins by running:
pipelight-plugin –list-enabled
Do you see Silverlight mentioned there?
The very first time that the pipelight plugin loads you should see a Wine dialog mentioning that SilverLight is being downloaded and installed.
Eric
Hi Eric,
yes, the pipelight-plugin commands runs ok, and I checked on my user home that .mozilla/plugin gets the correct link to silverlight library.
As with wine dialog mentionning that silverlight is being downloaded, well, I remember it showing up the first time I installed pipelight. As I had got problems, I did try to reinstall all over again, as I foun out I had to versions of the Alsa package, the converted one that I installed following your pipelight page instructions, and a previous one (an alien version, maybe from slackpkg+ with multilib settings). In fact I thought about cleaning out stuff in order to have wine downloading silverlight again to see if it fixes my problem, but I didnt know what I should erase.
thanks, Duodecimo.
Hi Eric,
Are you aware that your avidemux 2.6.x package creates a bin named ‘avidemux3’ but the avidemux.desktop entry still calls ‘avidemux2’? It’s not a big thing, but just in case you didn’t realized that.
And about your RSS. It’s great and I’ve subribed and use lots of your packages. Are you thoght on adding the link to the specific package in each RSS entry instead of linking to the whole repository? Just and idea.
Anyway, thanks for your work and happy New Year!
Hi, yes I am aware of that bug in avidemux which I did not spot at first (someone else pointed it out in another comment). I have a fixed SlackBuild but that won’t be used until there is a new version of avidemux probably.
I’ll give your RSS request some consideration.
Eric
Eric,
I have been spending the post couple of days recovering from something (that I probably did!) that erased the partition table from sda on my desktop machine. (At least it didn’t get my data drive. Anyway, when reloading calibre & all of its deps, I noticed that you have a “python-imaging” package & Slackware-current already has “pil”. Both seem to package the same digital libraries. I did not install your package, just to see what would happen. Calibre starts & performs simple tasks properly. Will I miss out on something more obscure by not having your package? If I do install your package, should I upgrade Slackwares package or put them in side by side?
Regards,
Bill
Yeah Slackware’s pil and my python-imaging packages are the same. I should remove it from the external dependency list.
Eric
Hi Eric,
i have installed the current slacky 14.1 and build a handbrake package with your script.
With the “Video Encoder” = x264 i get the error message “Segmentation fault”. So other encoder goes.
I read something about “Miscompilation with gcc 4.8”
Have you an idea for patch/workaround ?
Thanks, Björn
What if you use my package for handbrake instead of compiling it yourself? I can create H.264 videos on Slackware64-current without any segfault.
And what is “the current slacky”? The name of the distro is Slackware.
Eric
I compiling it self with your build files and get the error.
Now i found your package and have no problems with H.264. Thanks, Björn
Are you going to upload the 64 bit version of fbreader for -current? The folder is empty on both slackware.com & taper.alienbase.nl.
Regards,
Bill
Hi Bill
The package did not get uploaded for unknown reasons, but I will set that right tonight. Thanks for the notification.
Eric
Hi,
Just a small update regarding to XenServer Tools on Slackware. Finally I managed to get it working. I had to make the following changes:
– recompile the kernel with Xen guest support enabled in Linux guest support (it was a bit tricky to add it to lilo, because my device changed from /dev/sda to /dev/xvda)
– modify the distribution detection script in XenServer Tools to support detecting Slackware
– modify the guest parameters updater script, because “ifconfig” seems to have different output format on Slackware.
A future request for Slackware: Xen guest support enabled by default :}
Best regards,
kukukk
Hi Eric,
i am a brazilian user of slackware since longe time ago.
This days i try to figure out a way to start a diskless Slackware 14.1 client from a PXE + DHCP + NFS, without success. I can do this using a specific old 2.6 kernel, but in Slackware 14.1, i can’t figure out a way.
My PXE config sounds like:
DEFAULT pxelinux.cfg/vmlinuz-2.6 ip=dhcp root=/dev/nfs nfsroot=192.168.0.80:/tftpboot/nova-distro vga=791 init=/etc/rc.d/rc.S quiet rw
So if i change the old specific kernel (vmlinuz-2.6) to the new one Slackware 14.1 kernel (vmlinuz-huge-3.10.17) on PXE server, clients cant start due to a kernel panic error.
In Slackware 14.1 i will need a initrd?
Using the old “vmlinuz-2.6” i dont need any extra ramdisk, just the kernel.
Can you help me to solve this?
Thanks to you attention and sorry to my bad english.
Hey Eric,
Lookups like the slackware-current and slackware64-current directories on taper aren’t being updated.
Hi JKWood
Yes, a lock file was left (perhaps after a crashed mirror action) and it prevented further mirror actions.
I deleted the lockfile and the mirrors are syncing again.
Thanks for spotting and reporting!
Eric
Hello Eric,
is there a reason why your KDE packages don’t come with Kolab libraries anymore? I’ve tried and created build-scripts based on SlackBuilds-templates and your last build-scripts from 4.8 (or was it 4.9?) and got Kolab support for Kontact after compiling your KDE 4.12.3 packages. Additional dependencies are xerces-c (available on slackbuilds.org) and xsd (just created my own slackbuilds-script).
If you are interested in adding Kolab support again I could help if needed, otherwise I would simply submit my scripts to SlackBuilds. Personally I think it would be better if KDE would come with all needed dependencies, though.
Mike
Hi Mike
I wrote about how to re-add Kolab support here:
http://alien.slackbook.org/blog/kde-4-9-5-was-released-next-should-be-4-10/ which was well before Slackware adopted KDE 4.10. At that time, I discussed it with Patrick and we decided not to add the Kolab support packages to the distro.
You can still find the sources for libkolab and libkolabxml here: http://taper.alienbase.nl/mirrors/alien-kde/source/4.10.5/deps/ . A package for xerces-c is in my mail repository: http://www.slackware.com/~alien/slackbuilds/xerces-c/
Unless new dependencies have been added in the meantime, these three packages should allow you to add Kolab support by recompiling kdepim-runtime.
Eric
Hello! In multilib, I noticed that two packages have to be updated! openssl-1.0.1f and openssl-solibs-1.0.1f are vulnerables to Heartbleed. Many thanks for you excellent job.
Best regards,raffaele
Hi raffaele
Yes that was on the agenda for after work. By now, the updates should be in the multilib repository.
Eric
I liked Slackware to have requirements for Inkscape by default.
They are too much to install 🙂
I know it ain’t your fault but there is a error in LibreImpress that prevents the slides from showing unless hardware acceleration is turned off. It was quite alarming and I installed slackware on another machine and found the same problem with the latest update.
Appreciate all your work.
I am going to build a new machine and upgrade to the 64 bit Slackware. I suppose I should go with 14.1?
Ah, I live in Beijing so Chinese language support is useful for me. I got SCIM running about halfway- not very acceptable.
I fiddled with ibus for awhile some time back and finally couldnt make it perk. fcitx i got going for awhile but it was very unstable.
Any further thoughts? I am not going to try another Linux version. I am committed to being a slacker.
Hi Regnad
Yes, the issue with Impress has been known for sometime (http://www.linuxquestions.org/questions/slackware-14/impress-libreoffice-4-1-2-f5-freeze-on-slackware-current-4175480364/ for instance), nobody has fixed it yet unfortunately. Takes you by surprise.
Eric
Regnad, what are your issues with SCIM? I did not play around with ibus yet, so I can not talk about how (un)stable it is.
Eric
scim doesnt seem to work with some software like firefox. havent tried it with chromium though.
When I used scim it only seemed to be able to input chinese pinyin for qt4 applications. I then tried fcitx but it wasnt much better because it was erratic. I finally did a clean install to start over.
I am going to tear into the chinese thing full bore when I get time. Will probably go to Slackware 64 bit first though because I am wanting to build a i7 machine in a small portable itx size box for a project I am working on.
Any advice on Chinese input would be appreciated.
I installed the new KDE update, but now find that my audio does not work. Slackware finds the card and tests are ok but audio applications dont find the audio device.
Hi Eric. First of all, thanks for your *great* jobs for Slackware.
I’ running –current (64bit) with your multilib packages and last kde-4.13.0. On kde i’ve found that “baloo” ( the new semantic-indexing tool) eats lots of resources and eats lot of disk space ( for example, before i killed it, 7,9G of my homedir were filled by its stuff).
Removepkg-ed it, no chance to have dolpin so reinstallet baloo package but “chmod -x” its binaries.
Is there a chance to have dolpin not linked to baloo libraries?
Ah: my machine has a Amd-8core, 32G of ram and 5 hardisk ( my /home is on a raid-1 two 1G disks).
Thankyou again for your works…
HP
Hi Hubert,
Baloo libraries will remain linked to KDE binaries, but you can disable the indexing preocess in two ways: by removing the binaries “baloo_file_extractor” and “baloo_file_cleaner”.
OR, add your homedirectory to the list of excluded directories for indexing. If Baloo finds that you have excluded your $HOME, then it will disable itself. In “System Settings -> Desktop Search” add your home to the blacklist.
If you do not want the index-based search menu in dolphin, you can do the following:
$ cp /usr/share/autostart/baloo_file.desktop ~/.kde4/share/autostart/
$ echo “Hidden=True” >> ~/.kde4/share/autostart/baloo_file.desktop
There was a heated discussion on the KDE packagers mailing list about the bright idea of the developer, *not* to add a “disable indexing” button in the System Settings. I guess such a button will appear at some point in the future.
Eric
Hi Regnad
The updates to KDE should not have an influence on your audio… is there something else you have changed or updated recently?
Eric
Regnad, I would have to check using SCIM with pinyin input methods, to see how it affects Firefox. It used to work in the past, I have not paid attention to it recently.
Eric
TY alienbob…
Putting $HOME in the blacklist didn’t have effects on my machine, i dont’ know why.
And there is a bug ( as the developer writes in his blog) regarding Maildir directories ( i’ve such a dir, of course …) that can bring to a loop in indexing.
Anyway, “chmod -x” the binaries is good for me and i hope developers will put ASAP that damn button to stop their crappy software and that they will stop to think like Gnome’s developers do.
Have a funny WE
😉
HP
Hi Huber
I very much agree with your observation “stop to think like Gnome’s developers do”. The lack of the “disable indexing” button because the developer wanted to decide for the user that indexing is a “good thing” and therefore must be enabled always, strikes me as a worrisome example of how KDE software should _not_ work.
While the GNOME paradigm is “dumb down the interface, let the developer / UI designer make the decisions for the user” I have always though that the KDE philosophy was “if it can be tweaked, show a tweak button in the UI and leave the choice to the user”.
Eric
For anyone else experiencing failures to login to non kde sessions with kde-4.12.5 there’s a patch here http://sources.gentoo.org/cgi-bin/viewvc.cgi/gentoo-x86/kde-base/kdm/files/kdm-4.11.9-desktopnamescrash.patch?revision=1.1 .It works for me.
Big thanks to alienbob for the latest kde releases amongst many other things.
Hi Mickski56,
Well, the patch I used is https://projects.kde.org/projects/kde/kde-workspace/repository/revisions/e1724800ecf3c6a7035dfa7bcaa50b2a8f48688f/diff but it is the same issue that it fixes (the KDE patch is better than the Gentoo patch).
I was already compiling a new kde-worlspace package after this patch was announced on the kde-packagers mailing list and I am taking the opportunity to fulfill a request from Willy Sudiarto Raharjo to move the KDE sessions files to /usr/share/xsessions/ where XFCE installs its own session definition file already, so that other login managers can see these desktop session choices too.
Thanks for being alert.
Eric
hi – i am a longtime slack user. i was running 14.0 and using the ftdi_sio to talk to a printer and it was working fine, then i upgraded to 14.1 and the ftdi_sio now goes into an infinite connect/disconnect loop on the device. did 14.1 add something that is competing with ftdi_sio for the device? i can’t figure it out. the ftdi_sio source code hasn’t changed…
Hi eric spiering,
I have no idea – I do not use printers myself, perhaps it is wiser to open a thread on http://www.linuxquestions.org/questions/slackware-14/ and get a wider audience for your driver issue…
Eric
hello again, i have had this posted in LQ for several days now, but nobody has replied to it. i am thinking there is some configuration change in 14.1 that is now fighting for the ttyUSB0 connection, but i don’t know where to look to find it. do you see anything in these logs that might identify the culprit? This device connects fine to putty, so the device is good, but i need to use my minicom!
http://www.linuxquestions.org/questions/linux-software-2/ftdi_sio-connects-and-disconnects-forever-4175504613/
Hi eric
If this is a Slackware-specific issue, then you should request that the moderators move your post to the Slackware forum (http://www.linuxquestions.org/questions/slackware-14/) where there will be more knowledgeable people than in the generic forums.
The first line I read in your post, “I have slack 14.1, kernel 3.14.3” means that you are not running a stock Slackware 14.1 (which contains kernel 3.10.17 instead).
Eric
Hi Eric,
I just install your new chromium package and I have issues with the keyboard.
More details: I’m using a QWERTY keyboard with an IT layout and it’s impossible to type every accented letter:
à is ‘
è is [
ì is =
ò is ;
ù is nothing.
It’s a chromium issue because everywhere in the system typing is good and I tested this on two 64bit slackware one with kernel 3.10.17 and one with 3.14.1.
With the previous version of chromium all was good.
What would suggest to do?
Hi Alberto
Are you perhaps using SCIM or Ibus as text input methods? There is a bug in Chromium 35 apparently with XIM, as found here: https://code.google.com/p/chromium/issues/detail?id=360388
Eric
Nope.
But I found a similar issue (probably the same, but explained better than mine) and setting the LANG variable before starting chromiu seems to help:
$ LANG=it_IT.UTF-8 chromium
In other words, browser keyboard layout is “locked” to en_US.
https://code.google.com/p/chromium/issues/list/detail?id=376246
Sorry: https://code.google.com/p/chromium/issues/detail?id=376246
Upgraded just know: no problems at all with “italian” letters ( or keyboard).
I’m on 64–current with the entire Aliens’ stuff (multilib, kde, eccecc).
I have:
export LOCALE=it_IT.UTF-8
export LANG=it_IT.UTF-8
in /etc/profile.d/lang.sh
italian keyboard in Kde and:
Option “XkbLayout” “it”
in the InputDevice section of my /etc/X11/xorg.conf
HP
It doesn’t work for me.
Alberto: what family of fonts are set in chromium?
I guess standard ones. Anyway, after rebooting everything works fine. Thanks for your suggestion =)
Sadly there’s another problem: after updating, chromium doesn’t see java.
Worst release ever =(
P.S. In firefox everything works fine.
Hi Eric
Tried updating weekly version of calibre, but according to slackpkg+ there are no updates. Checked around various mirror sites and it seems that only the master has been updated.
Hi Alberto
Starting with chrome and chromium 35, Google has removed support for NPAPI plugins (Netscape Plugin API i.e. the mozilla browser compatible plugins). This means that things like the icedtea-web Java plugin stopped being supported, but also the Pipelight plugin and several others that do not “talk” Google’s own PPAPI (Pepper Plugin API).
It sucks.
Eric
Hi Phil
If everybody mirrors from my “taper” mirror then nobody will have had new packages. It looks like the mirror cron job had become stuck.
It’s been repaired now.
Eric
Hi Eric,
thanks for your reply.
I don’t use openJDK, but Oracle’s official jdk (repackaged with the slackbuild included in Slackware).
Does it make any difference?
Cheers Eric
My mirror now up to date 🙂
Hi Alberto
The closed-source Oracle Java browser plugin uses the same NPAPI protocol as the opensource icedtea-web browser plugin. Both are incompatible with Chrome and Chromium 35 and higher.
Eric
I see =/
The worst part is that also google talk plugin is blocked now.
Alberto, the newest version of the HangoutPlugin should be a PPAPI version and work in Chrome/Chromium 35.
Eric
I tried with the last one (version 5.4.1.0 – according to what the SBo slackbuild extracted) and it is not visible among plugins =(
If you succeed in making it work, let me know =)
Thanks again!
Hi Alberto
The google-talkplugin 5.4.1.0 installs both the NPAPI and the PPAPI libraries. It takes a small change to make the PPAPI library available in chromium if you are using the SBo script to create a package.
Below is the patch I applied to the google-talkplugin.SlackBuild :
Eric
# —–8< ---------------------------------- 26c26 < VERSION=${VERSION:-4.9.1.0} --- > VERSION=${VERSION:-5.4.1.0}
37a38
>
42a44
> # Just in case:
65,66c67,75
< chmod 0755 $PKG # Put this back. < rm -rf etc/ # The cron job is debian/ubuntu only. --- > # Put this back.
> chmod 0755 $PKG
>
> # change /usr/lib/chromium-browser to /usr/lib/chromium
> mv ${PKG}/usr/lib/chromium-browser ${PKG}/usr/lib/chromium
>
> # Remove cron update script:
> rm -rf ${PKG}/opt/google/talkplugin/cron
> rm -rf ${PKG}/etc
# —–8< ----------------------------------
It works!
Thanks a lot, Eric.
Hi Eric,
since Chromium 35 doesn’t support NPAPI plugins any more, could you make available again your last Chromium 34 package?
Thanks a lot!!
I get an error when I try ann open a book on the command line with fbreader 0.99.4.
Gentoo had the following patch that works for me:
— fbreader-0.99.4-pristine/zlibrary/ui/src/qt4/filesystem/ZLQtFSManager.cpp
+++ fbreader-0.99.4/zlibrary/ui/src/qt4/filesystem/ZLQtFSManager.cpp
@@ -49,3 +49,4 @@
path = replacement + path.substr(1);
– }
+ }
+ ZLUnixFSManager::normalizeRealPath(path);
Hi Brad
Ah, weird! I see the same crash, and indeed only happens when you try to open a book on the commandline.
I am going to look at that patch, thanks.
Eric
Hi JesusM
Chromium 35 is a security release, so it is not advisable to keep running 34.
If you need Java or Netflix, then Firefox will be your only option.
If you want to know how hard it is to switch pipelight to a PPAPI plugin because of the sandboxing of Chrome plugins, read https://bugs.launchpad.net/pipelight/+bug/1321505
Eric
Hi Eric! I’d like to thank you for your build of Chromium 35. It feels much lighter and faster than the previous versions.
However, there is a problem and I’d like to troubleshoot it in order to know whether the issue lies in the build, or the software itself, or my own configuration.
My keyboard layout is US International, which has dead keys. Now, starting from Chromium 35 I cannot produce some characters with dead keys, but I can produce others. Among the characters that I can write are: á é ñ, and among the characters that I cannot produce are: « » ¿ ç. Some of them (like ¿) are essential to write proper Spanish.
Thus: Is this due to some ./configure parameter? Is this because the software itself (I understand they’re using a new toolkit).
My own configuration may also be the culprit, but I doubt it because I had no problems until Chromium 34.
Thanks again,
Eduardo
Some comments:
1. I got SCIM to work in chinese and it works very well indeed and I have a fully functional bilingual (for me) system.
2. The latest version of Libre Office is quite good and I am transitioning completely away from microsoft office except for shared edits with stored changes (just because I dont feel completely comfortable with this feature in microsoft office.)
I still have to disable hardware acceleration manually otherwise Libre Impress hangs on slide shows.
3. I recently had a horrid experience with adding a 3rd hard drive to my machine which has to be added to an SATA slot number smaller than the boot drive. Finally i got it going using UUID and mkinitrd and lilo.conf editing. This really was a pain and took a lot of time to figure out and I almost despaired of adding the 3rd hdd. The documentation on this was not so clear and I had to dig. Others might not be so determined as me.
Just to notice you the new release of mirall
;-)))
Thanks in advance!!!
Hi Eric
Please I need chromium-33.0.1750.152-i486.txz you have this package? Please post link!
Hi Eric, I noticed you upgraded chromium to 36. I use Slack64-current multilib. I’m thinking about switching to firefox because since the latest X.org upgrade in current, chromium in Slack64-current began to be really sluggish when typing text. This began to be apparent in chromium 35 (after the X.org upgrade) and continued in chromium 36. To this add the known issue with dead keys and foreign languages.
I’ll continue to upgrade chromium as you provides releases, but for now it will become my second browser, not the default. The typing issue makes it very uncomfortable to use it.
Thanks again for all your effort. I hope this gets solved soon.
Regards,
Eduardo
Hi Eric,
as always great work with the new chromium release.
I’d like to report a bug that wasn’t there in previous releases: it is not possible to set it as default browser.
I don’t know if it is due to the build process or to chromium itself.
Hi Alberto,
Works here with Chromium 36.0.1985.125 and running Slackware (almost-)current and KDE 4.13.1.
Note that setting it as the default browser changes the xdg-open behaviour.
How -current does -current need to be? I am running multilib -current but havent updated to the most recent -current. I notice that you are also not-quite-current. The latest KDE and chromium are up and working ok. I find some problem with KDE occasionally (kwin blows up and gives and error message but I can keep working) and LibreCalc does weird stuff sometimes but it’s all tolerable.
Actually I am a bit leery of the latest -current with the display problems.
Hi Eric, I’m happy to say that Chromium behaved itself again after the latest X.org update (I use Intel chips). The issue with international characters continue, though. Thanks for everything!
Hi,
Thank you for providing this site and your slackware packages.
My apologies if this should be posted to the LibreOffice dev group. I searched for a ticket on this but couldn’t find one.
I have been using LibreOffice from your slackbuilds (slackware64-14.1) and have had no issues until the 4.3.0 release. Just recently, I tried saving in MS *.xls and *.xlsx formats and received an error from LibreOffice stating that saving failed. The export works with OpenOffice 4 and previous versions of LibreOffice.
Just wondered if you or anyone else has had this issue.
Cheers,
Fred
i regularly save in those formats .xls .xlsx from libreoffice calc and dont have any trouble.
i am running slackware 64 -current (linux 3.14.16) and the latest libre office. i had some issues with calc blowing up after font changes a couple versions back but it works fine now.
Hi Eric,
I just installed the last version of your pepperflash package (15.0.0.152) and I noticed a weird thing.
In chrome://plugins page I have the following flash version (but the path is the pepper’s one): http://imgur.com/lXLPNzk
I’m using chromium 37.0.2062.94.
Is this normal?
There is nothing wrong with the way the plugin reports its version to web sites. Just check http://www.adobe.com/software/flash/about/
The problem is in the /etc/default/chromium file where I try to determine the flash version. The string in the binary changed so it’s reporting nothing now, and in chroime://plugins that translates to this weird 11..2.999.999 version.
Make sure the version determination string in /etc/default/chromium is changed to:
flashversion=$(strings $flashso|grep “LNX “|sed -e “s/.*LNX //”|sed -e “s/,/./g”)
It’s really only cosmetics, but I refreshed my package anyway to fix this.
Hi Eric, just wanted to ask you if you could update your scripts regarding SimpleScreenRecording. The changes are quite important imo.
Building from git works pretty well, the “simple-build-and-install”+”postinstall” scripts do all the magic (just needed to disable pulseaudio).
Sorry I couldn’t send you proper patches for such an easy task to save you the minutes, I lack knowledge on the topic.
Hi Jerónimo – what scripts should I update and why? And with what/
Eric
The slackbuild from here: http://www.slackware.com/~alien/slackbuilds/simplescreenrecorder/build/
Because it does everything flawless but I get no exec file (notice that it download latest git master when it check the current date). Can’t find the program itself anywhere.
I guess with the “simple-build-and-install” script (diff).
Sorry If it’s me and I’m doing something wrong.
hi Eric, three (four?) years ago I downloaded a package of yours — Shisen — built without all the (KDE) cruft. I used to love playing it and, having a new 2nd-hand computer, looked to download the game again. alas, cannot find it onsite, search returned no useful results. can you help? thank you.
Hi jr
I never had a “shisen” package separately in my repository. Shisen is part of KDE, can not be shipped separately.
Perhaps you are thinking about another Mahjongg game? In my repository I have xmahjongg (quite old, just a 32-bit Slackware 12.1 package).
hi Eric, thanks. I played kShisen a few days ago (love these tile matching games) and must have got mixed-up. I’d appreciate an url for the mahjonng game, it’s the one with the flowery(?) background image, right? thank you in advance.
regards, jr.
Hi Eric,
few week ago you talked about the 2.x release of Calibre and the doubts you had about embedding Qt5 libraries or keep them separate. In the meanwhile your Calibre package would remain un-updated. Have you made a choice?
Thanks!
Hi JesusM
I am too busy with work (the kind that earns money) to spend a lot of time on Slackware at this moment. I have decided not to try and build a calibre 2 package for the moment because it needs research time which I do not have right now.
Hi Eric,
Thanks for your packages.
I’ve tried your Chromium package on my Slackware-14.0 (release: 37.0.2062.94).
But I noticed the omnibox doesn’t work as usual: if I put there, for example “hello world”… well nothing happens!
I expected to be redirect to:
“w w w . g o o g l e . c o m/search?q=hello world”.
I’ve tried to change search engine and noticed that the matter is in “%s” variabile not expanded regularly.
For example if I create a new serch engine called with following string:
myengineaddress/search?q=%s
Then if I put something in omnibox and press enter, appears the address:
“myengineaddress/search?q=”
…without “%s” expanded.
Now, default search engine string (related to google engine) is made of just variabiles:
“{google:baseURL}search?q=%s{google:originalQueryForSuggestion}{google:assistedQueryStats}{google:searchFieldtrialParameter}{google:bookmarkBarPinned}{google:searchClient}{google:sourceId}{google:instantExtendedEnabledParameter}{google:omniboxStartMarginParameter}ie={inputEncoding}”
So here the answer to why nothing happened when I tried to search something using omnibox.
Can you confirm this release behavior?
Or pheraps could be a my local system matter?
Are you going to update chromium package also for slackware 14.0?
Thanks in advance! 🙂
Since I recently bought a new laptop I had to upgrade to Slackware current just to get it to boot. Since I did that I thought I’d try your kde upgrade to 4.14.2. Worked fine except for Okular which could no longer read mobipocket format. I had to back out to Slack 14.1 to get it working.
Just to not only complain here… I really really really find your multi-libraries handy. Thank you for that hard work and saving me from the aggravation. 🙂
I looked at your remark about Okular and I have found the culprit.
There is an issue with the order in which some packages are built.
I have fixed that order and the next KDE 4.14 release will have an Okular with mobipocket support.
In the meantime, you could install calibre or FBreader which both support .mobi.
Hi,
installed package wireshark on my slackware but after installation when I try to run the following error occurs:
bash-4.2 # wireshark
wireshark: error while loading shared libraries: libgnutls.so.26: can not open shared object file: No such file or directory
Can you help me?
Thank you
You probably installed my wireshark package on Slackware 14.1 which I built for Slackware 13.37. The glutls package was updated since Slackware 13.37 and the package needs a recompilation.
You can download the wireshark “build” directory and run the wireshark.SlackBuild script. The package should be available in /tmp when the compilation finished.
A request. In a future update of ffmpeg package would it be possible to add libfdk_aac to the encoders? And thank you very much for your hard work. Obrigado!!
Hi fabio, I can do that.
hi eric,
i wanted to ask u this question quiet longtime. I have a lenovo t400 and i can recall u use to have the same! … i have always run SC on it, and everything was quiet fine, minor problems but they could be easily fixed … since pat updated “x/*” on July 15, which messed up almost everything. Despite all my efforts i couldn’t get the issue with graphic fixed … openGL got massive problems, the lcd is partially colored with blue or black till refreshing opened windows, VSync cusses freezing and etc … i just wanna know if u have had also the same problem on ur t400? any solution to this? (of course if u still have ur t400!) … and do u know any way to upgrade Slackware Current to the version before July 15? (Just to mentioned i don’t have any of those problem on S14.1)
thank u in advance,
bests
Hi Inman
I still have that T400 and use it daily – it is my main workstation for work and home.
I am running Slackware64-current on it and have never experienced the problems you are describing. Could it be that you missed some new packages that you forgot to install, or else have not deleted some package that was removed in Slackware-current? A properly configured slackpkg will offer to install all missing packages when you run “slackpkg update ; slackpkg install-new”.
By the way, “slackware-current” does not have backup versions. You either have what’s available today, or you have to go back to Slackware 14.1 which is the most recent stable release available.
Thanks for the chromium src scripts. Just a small thing to say. I don’t use printing and so cups is not installed. The chromium build fails when it tries to do the gyp stuff and after a lot of digging I found that I needed to add -Duse_cups=0 and then all worked. I don’t enable NaCL (sodium chloride I use) and I don’t even know what advantage there is using NaCL (other than food tastes better with it, but might be cooks fault and not food’s fault).
When you use one of my scripts to compile a package, I only guarantee its working on a full installation of Slackware, since that is how I build them myself.
NaCl is Google’s Native Client. See https://developer.chrome.com/native-client so you can decide for yourself if this is useful for you.
I understand the full-installation aspect of Slackware and it is a key thing for anyone using Slackware to understand, especially newcomers. I mentioned the cups config in case someone else might stumble into it. Thanks again.
(FWIW I don’t need the NaCL)
Just noticed the update on ffmpeg. Thank for the new decoders/encoders. I know you have a real life, and taking the time to include requests was very kind of you.
Hi Eric,
RE; ip-address proxying
This was posted from the Mistress of the house’s computer
and if this got through, your explanation of my ISP proxying traffic from my computer through somewhere else, is verified.
Will expand further off-line from your feedback-of-blog.
Regards,
mike.
Thanks so much for the great works that you have done. I have been redirected countless times from searches to your works and every time they work!
Thanks again!
Ed
Hi Eric,
First of all I have to say that you’re one of the main forces that help me to stay and develop using Slackware. I’m running current-64, multilib and a lot of alien packages absolutely updated. I start using qemu and vde networking from your first packages. My present issue is on vde. For some reason rc.vdenetwork (nat option) could not start dnsmasq and serve my qemu VMs. The error message is: “dnsmasq: failed to create listening socket for 10.111.111.254: address already in use”. Could you give me some hint ?
Cheers
Gilcio
Hi Gilcio
It sounds as if you already have a DHCP/DNS server running on your computer.
You’re right. I did a ps ax | grep dhcp and results: “6476 ? S 0:00 /sbin/dhcpcd -B -K -L -G -c /usr/libexec/nm-dhcp-client.action -h orion eth1” but I have no idea of what program could have started it?. I will look at it in more detail. As always thanks for the fast answer in a sunday afternoon.
How do I make these stop? I get emails but there’s no info on how to make them stop. If I subscribed I made a mistake.
Gilcio – the dhcpcd is the “DHCP client daemon” which manages your computer’s IP address. It is not a DHCP server.
If there is a DHCP server, it will be listening on port 67 and you should find its process with the command (as root): “netstat -panel | grep :67”
R.H. – all the emails you receive from this blog are because you checked a subscription checkbox below one or more articles. I do not send them myself.
The emails you are getting should contain a link that allows you to manage your subscriptions.
I figured out what’s happening. Some emails from your blog show the unsubscribe and some do not.
emails with subject [New comment] do not show the unsubscribe and those with [New post] do.
I can see the unsubscribe link always:
http://i993.photobucket.com/albums/af60/erpopo/AlienPost_zpsc252602b.jpg
Eric, sorry for a late acknowledge of your answer. I did a fresh reinstall of slack64 current and all my problems disappear. Thanks for your time
Eric,
When are you planning to upload patched (CVE-2015-0235) glibc-multilib?
Will the updates on glibc (both stable and current) trigger an update in the multilib set? yes, it is a request. thanks Eric.
Yes, please update multilib glibc as soon as you can. It’s a pretty dangerous vulnerability. More details on this and other up-to-date computer vulnerability news:
http://threatpost.com/of-ghost-glibc-vulnerability-patching-and-exploits/110719
Thanks for making multilib! I couldn’t use more than half of what’s out there, if there weren’t this easy multilib setup! 🙂
You guys don’t seem to realize that I do all this Slackware stuff in my spare time. If you start demanding that I must do things I will start demanding money in return.
The updates will arrive as soon as I have time at home to build them.
in my defense: 1 – You created the space for requests and 2 i asked politely, not demanded. 3 if i had money you wouldn’t have to demand. It would be my pleasure due to the great work you do for us all. sorry to bother man, but you are always so fast that we grew spoiled hehe.
Requets for new stuff is one thing. Requests to work faster is not acceptible to me. Yes, you were polite about it, fabio, and it was on the planning list anyway. My comment was more directed at Alex and MajorLunaC – I am well aware of glibc updates. This GHOST vulnerability is hard to exploit, actually there are no real-life exploits yet (apart from one for Exim which Slackware does not ship). You should try not to be so paranoid. And if you are, you know where the source code is.
Wow. So the words “please”, “as soon as you can”, and “Thanks” now mean “Work faster slave! Right now, damit!” in modern internet lingo? Let me interpret for those who think everyone is demanding instant results for a completely free repository made in spare time, and that everyone is ready to take a bite out of you. Maybe we use different dictionaries:
“please” = “I beg of you”. “I ask of you kindly”. Even “I seek your help”.
“as soon as you can” = “Whenever you are willing and able”. “I now you work hard, and you do in your spare luxury time in which you could be having fun, so if you happen to find the time, anytime in the distant future”. “Whenever you see fit”.
“Thanks” = “I am in your debt”. “I really appreciate it”. “I don’t think I could have done it without you”.
**As for your comments:
“I am well aware of glibc updates”: First I’ve heard you mention it.
“it was on the planning list anyway”: I would love to see that list. If I had known you knew of it and it was set to be updated, I would never have posted except a “Thank You for such a quick fix! Nice Work! You work too hard! Take it easy!”
“actually there are no real-life exploits yet”: A vulnerability is a vulnerability, no matter how small it is. In my opinion, vulnerabilities get their own level of priority far above anything else. Disclosure of the vulnerability increases the chance that someone will try to exploit it 10 fold or more, which is why vulnerabilities are usually disclosed privately, quite some time before the public disclosure, to those who can and do fix them. Once the vulnerability is publicly popularized, new ways to exploit it are often explored.
I need to come up with something to say to those who consider me paranoid, like “Wait till you see what I got off your computer …”
now that i saw the guy before me had asked the same, i fell quite stupid for posting before reading everything. As for the vulnerability itself, my primary concern was about getting the applications that rely on multilib working again, since few mortals have the solid knowledge in programming to really exploit it. Thanks for the updates and as for me, will wait quietly next time.
MajorLunaC, and a good day to you too.
Actually, no one gets to see my TODO list. And if you would go looking at historical data of the multilib ChangeLog.txt you would notice that I am usually only hours or a day behind on Slackware proper. However, these were glib updates for 13.37, 14.0 and 14.1 combined with the KDE5 which cost me six weeks of multile hours per day work. At some point there is a decision to make about the balance of hobby work (which this is) and tending to family. The multilib glibc updates would come, if you were doubting that, you are quite new to Slackware and this community.
As for vulnerabilities: they are real but overrated. A vulnerability must be exploitable to be a vulnerability. And exploitable does not mean that it is exploitable _on Slackware_
Nevertheless, if it bothers you (and I respect your point of view) there is always the option to go back to Slackware’s patched glibc – temporarily – until the time that I release new multilib versions. All you lose is the capability to run 32-bit programs. You gain peace of mind by knowing your system is patched against a vulnerability.
Finally: you do not get so say what I have to do, unless you are my employer. And when you do, even if accompanied with please and thank you – you can rub me the wrong way if I just came home from a bad day at work and don’t need more people on my back.
End of statement.
Quick suggestion for when you get the time (if you feel like doing it, of course). A bump on LXQT to the 0.9 version. And thanks for the 0.7 version. Excellent for “reviving” old machines with slackware.
Hi fabio
Building LxQt 0.9 is somewhere on my TODO. Their migration to Qt 5 and (some) KDE Frameworks made the update non-trivial and I wanted a stable KDE 5 before looking closer at LxQt.
First of all a heartfelt thanks for your sterling work producing reliable slackware packages 🙂
However, your latest vlc package (2.2.0) doesn’t handle HD (720p and above) as well as the older 2.1.5 version: video stutters or freezes.
I downgraded to your 2.1.4 package (in the slackware-14.0 repo) and everything was back to normal.
I know that building vlc packages is a PITA. It could also be that my slightly outdated hardware could cause the malfunctioning. Still, if you can find the time to look into it, it would be highly appreciated.
Hi KG
Start “vlc -vv” (or with one v, or with three, depending on the amount of debug info you need) and see if there are any hardware acceleration related errors when you play a video.
I have no problems with playback of HD video here.
Hi Eric and thanks for a swift reply!
Embarrasing: after a new upgrade, the videoclip that caused trouble yesterday ran OK…
vlc -vv creates a huge amount of info, and repeated entries like this:
“[b5487e98] avcodec decoder warning: More than 4 late frames, dropping frame
[h264 @ 0xb549cf40] Frame num gap 19 17”
Maybe it should be put down to hardware. I’m on an old netbook with a Celeron processor, and when playing the HD clip in question the CPU load is as high as 80%. No wonder if there’s a bit of stuttering.
So for now I’ll stick to the latest version 🙂
diff for new version of google-chrome.SlackBuild
77c77
ar p $CWD/google-chrome-${RELEASE}_current_${DEBARCH}.deb data.tar.lzma | lzma -d | tar xv || exit 1
Hi Bryan
The recent debian packages are using xz compresson instead of lzma compression. The file is now called data.tar.xz.
The google-chrome.SlackBuild in slackware-current was updated for this, early March 2015.
Is there any chance that the latest “slackware-current” might no longer work with the files you have in the “current” directory of the multilib feed? I have been doing multilib upgrades with your packages for quite a while, but now when I run everything–with a fresh download and the same commands as before, glibc seems to get hosed, and the system is then unusable. I’ve tried twice, both times with the same result–can’t get at any commands and have to start the install over.
Hi Brian
The multilib packages in the “current” directory of my repository are the ones for slackware-current as you can download it from any Slackware mirror. Check the package versions and BUILD numbers, they should match the versions of Slackware64-current (except my packages have the “alien” tag added to the BUILD number).
Can you double-check that you are indeed downloading the right packages, and that they have not been corrupted?
If you can’t find what’s wrong perhaps you can share:
– where you are downloading those packages
– what their md5sums are
– whether the gpg check is OK
– what commands you are using to upgrade glibc and in which order you issue these commands.
Hi
Mirror contain “deleted” package after huge update (20150421)
eg:
http://taper.alienbase.nl/mirrors/slackware/slackware64-current/slackware64/k/
contain kernel source 4.14.33 and 4.18.11
same kde 4.10.5 amd 4.14.x
same type of thing in the iso
http://taper.alienbase.nl/mirrors/slackware/slackware64-current-iso/
Hi David
The primary mirror which I use is http://slackware.osuosl.org/slackware64-current/source and its contents are correct.
However the rsync URI which is rsync.osuosl.org::slackware/slackware64-current/ does not offer the same content: it still shows the old packages which should have been removed.
I use the rsync URI which is why my taper mirror shows all these redundant packages.
I am sure that this is an initial sync issue and that the primary mirror will be sanitized soon. Then taper’s contents will go back to normal too.
Good Afternoon,
Early this morning, my time, I went to download your latest -current iso and noticed it was 4.3 gigs in size.
Is that correct? I had thought the size of the iso had been reduced to somewhere over 2 gigs?
Thanks.
cwizardone, it takes time to properly sync the mirrors. And if you sync at the same time it’s possible to get the old AND the new packages all together…
My taoer mirror had that issue, and as a result that ISO image also contained all the original as well as the updated and new packages, causing the filesize to explode
The mirror has been re-synced and the ISO re-created. It is now a proper 2.6 GB in size.
1st off, thanks alienbob.
Since we’re on the topic of multilib is there an https location for multilib files? I did look around and try some urls but none worked. I wonder if osuosl or kernel.org would or could do that hosting.
hey also I see you’re a sci-fi reader & collector, I have some and would be willing to donate one to you if interested. I have pohl, poul, lem, asimov, et.al. I could ship it whatever’s cheapest i can find to anywhere that usps goes I guess.
Hi R.H.
Currently there is no https:// URL for download of multilib files that I am aware of. Downloading packages over HTTPS only adds strain to the server because it has to encrypt all the data. It adds no security… if you want to have security you can use the GPG signature files to verify your downloads.
Thanks for your offer to donate a Sci-Fi book but I think it will cost you too much in shipping. I have a wall full of books and no room anymore 😉 I switched to reading e-books because that is space-saving.
Just to inform, Libreoffice won’t start after the last batch upgrades on -current.
symbol lookup error: /usr/lib64/libreoffice/program/libi18nutil.so: undefined symbol
as well as other libs missing. Thanks for your work on the updates!
Hi fabio, I know. I have new (and working!) 64-bit LibreOffice packages ready and the 32-bit package is still compiling (will take the rest of the day while I am at work).
Your speed is amazing man!!! libreoffice working perfectly. I know it’s minor and the focus now is kde5 but just to inform, in kde settings on -current it’s impossible to get new themes and window decorations, by clicking the option an API error occurs. Thanks again for the lightning fast libreoffice packages.
Note on my email of April 18:
I just fetched a new slackware-current ISO over the weekend, and using your guide’s lftp URL things worked perfectly. So it must have been some kind of hysteresis in the two feeds at the time I was doing my previous work.
Thanks for your wonderful work!
about kde themes not working, if someone got the same. Just delete .kde folder in your home. worked for me. sorry about the noise early.
Dear Alien,
Thank you for all your great packages and work on Slackware. I have two questions about some of the packages you host:
First one is regarding Chromium. I use Chromium mainly for Netflix and some other small audio related things (works very well with ALSA emulation), but I’ve been unable to use it for a month or two now due to an unexplained “Segmentation Fault” on any webpage load. I’ve tried several different new versions, deleted all Chromium config files, and even with an upgraded kernel (I’m now on 3.18.11) it still segfaults. Any idea what could be the cause?
Second question I have is regarding the multilib versions of Mesa. I recently have been trying out the nouveau drivers because they bring a lot of nice features that the nVidia drivers don’t and in combination with the new kernel, they now work great with my laptop. However, the problem is that despite the 64 bit versions of Mesa packages for Slackware are OpenGL 3.0, the 32-bit packages spit out OpenGL 1.4. This is very annoying considering that Source games can’t even run with such an old version, and Wine always uses the 32 bit version of Mesa even in 64 bit prefixes. Is there anyway to get the 32 bit libraries to utilize OpenGL 3.0?
Appreciate any help or response. Thanks.
Hi Xylemon
You don’t tell what version of Slackware you are running, and on what type of hardware. If Chromium segfaults for you, perhaps you are able to attach a debugger to the session to get a backttrace? Or else, try adding the commandline parameter “–bwsi” to Chromium (“browse without sign in” which is Guest Mode with disabled extensions). If that makes Chromium run again, it may be one of your extensions which is causing the crash.
About your OpenGL question – you do not provide any example commands that you’ve tried inclusing literal copy/paste of the command output. I can not start guessing at what’s happening.
Thanks for the reply,
I am on Slackware 14.1 64-bit with Kernel 3.18.11 and I generally use the current packages with your multilib repo. I have a GT 755M card, and this is the rest of my system specs: http://shop.lenovo.com/us/en/laptops/lenovo/y-series/y510p/#tab-tech_specs
Regarding Chromium, I have tried removing any type of configuration file and install. Trying it with “-bwsi” has no effect. I’ll look into a backtrace or debugging soon. If it’s worth anything, I think the problem started when that version released which supposedly did not work on 3.10 kernels or something of the sort.
On the Mesa/OpenGL topic, the info I get regarding the 32 bit version of Mesa is this:
/usr/bin/32/glxinfo | grep “OpenGL”
Which in return spits out:
libGL error: unable to load driver: i965_dri.so
libGL error: driver pointer missing
libGL error: failed to load driver: i965
libGL error: unable to load driver: swrast_dri.so
libGL error: failed to load driver: swrast
OpenGL vendor string: Intel Open Source Technology Center
OpenGL renderer string: Mesa DRI Intel(R) Haswell Mobile
OpenGL version string: 1.4 (3.0 Mesa 10.5.3)
OpenGL extensions:
From what I understand, I should see 3.0 for the OpenGL version just like what the 64 bit version of glxinfo spits out:
OpenGL vendor string: Intel Open Source Technology Center
OpenGL renderer string: Mesa DRI Intel(R) Haswell Mobile
OpenGL core profile version string: 3.3 (Core Profile) Mesa 10.5.3
OpenGL core profile shading language version string: 3.30
OpenGL core profile context flags: (none)
OpenGL core profile profile mask: core profile
OpenGL core profile extensions:
OpenGL version string: 3.0 Mesa 10.5.3
OpenGL shading language version string: 1.30
OpenGL context flags: (none)
OpenGL extensions:
OpenGL ES profile version string: OpenGL ES 3.0 Mesa 10.5.3
OpenGL ES profile shading language version string: OpenGL ES GLSL ES 3.00
OpenGL ES profile extensions:
Hope this is more insightful than my last comment.
Hi Eric,
surely it is my mistake, but I dont’ see where …
you have announced Calibre 2x for both Slackware 14.1 and -current. Indeed, today I saw at the RSS channel the notice for Calibre 2.30.
But when I try to update using ‘slackpkg update’ I still see Calibre 1.48.
I did update to Calibre 2.28 by downloading directly your pack from the repository, not using ‘slackpkg’.
What am I doing wrong?
Thanks once mor time for your job!
Hi JesusM,
I guess it is an issue with slackpkg, because the newer calibre package is there in the 14.1 repository, next to the 1.x version of the package.
To end the confusion, I will rename the calibre-1 package for Slackware 14.1 to “calibre1” like I did in the slackware-current repository.
Howdy AlienBob
Sorry for posting my question on “Parental control on the Linux desktop”…I misunderstood you statement above(“This is not meant to be a replacement for…). I stand corrected.
To the point. I often worry about the timeliness of many articles seen on the web with the fast pace of change that we live in. Your post on creating a safety net for our children is wonderful and meets my expectations quite well. Wishing to avoid a false sense of security, however, before implementing your suggestions I was wondering if there is anything you would change or add six years later, if that is in fact when you wrote the post.
So, any comments, changes, suggestions?
Hi senecum,
My son grew up in those six years so I am no longer using parental control, but the concept is unchanged and the software you can use is also still the same.
The most recent release of tinyproxy is four years ago, but a lack of new releases does not mean that the sofware has become unusable or unfit. Some programs work well enough, especially if they are small and focused on one task, that they do not require a continuation of upgrades. Just comple the most recent version of tinyproxy for your computer and try it out.
Squid is the bigger brother of tinyproxy and of course you can consider that as an alternive – it’s just a bit more complex to setup.
Dansguardian or squidguard are the programs that do the actual content checking and filtering, those are still current.
If you implement this solution yourself and find that the article needs a refresher or requires fixing, please let me know so we can discuss it; or leave your comments in the “talk” section of the Wiki page. Credit where credit is due of course.
Thank you, AlienBob – I would say Eric but I like AlienBob too much… every time I see the name the tune from Sponge Bob invades my senses, I get “happy feet” and brings a smile to my face, a side effect of having small kids :?)
I will let you know if I come across any issues that need your attention.
Best
Howdy, again AlienBob
In the spirit of trying to give something back, unable to offer anything else of value, I want to share this great little gem I came across today (the performance, not the song itself, which is well known already). I hope it will be of your liking and that it will shed some sunlight on those “rainy days”, or after one of those “wonderful” days at work…
Best
P.S. the blog includes “life and everything” so I hope this post is not terribly inappropriate…delete at your discretion if it is.
Hi Eric,
I am very impressed, that you have mastered integrating the WideVine-Plugin into the Open-Source Chromium browser, and even to playback Netflix videos!!
So, I am trying to do the same for my Raspberry Pi (with Raspbian OS). The official Chrome browser does not work there, because it is not built for ARM architecture.
Chromium does work, but without the EME extensions.
So, do you have any tipps for me how to get it working on the Raspi, or even could provide a ARM build of your WideVine plugin?
Would be greatly appreciated 😉
Cheers, Chris
Hi Chris
I do not think that’s going to work. I do not compile the Widevine CDM library, becuuse that is a piece of closed-source technology. What I do is compile the “adapter library” which interfaces Chromium with the Widevine Content Decryption Module. The Widevine CDM itself is then extracted from the official Google Chrome RPM binary.
For an ARM platform, you will require a binary Widevine CDM library which is made available by Google.
ChromeOS is available for ARM (my Samsung Chromebook runs it) so there *is* a version of Chrome that works on ARM. I have no idea (never tested) if Netflix works on the Chromebook, but if it works it might be possible to extract the Widevine libraries from the ARM archive of Chrome.
Thanks for your help!
I found some useful information on how to extract the Widevine plugin for ARM architecture here:
https://stackoverflow.com/questions/30732428/extract-files-from-chrome-os-chromebook-recovery-image
But what to do next??
– Which binary files are needed exactly?
– How to include them into the Chromium ARM build? (Therefore I would need your “adapter library”, right?
Hi Eric,
I would like to report that something went wrong with your latest Chromium release (the one without the OK Google binary blob, v. 43.0.2357.125-x86_64-2alien)
Chromium suddenly appeared big and blurry, as if I were using a 800×600 screen. Totally ugly.
Only after launching Chromium with the ‘–force-device-scale-factor=1’ gives me a normal appearance.
For the record, I am not enlarging or forcing any scaling on the display at all. All previous Chromium versions looked fine.
Thank you again for all your efforts.
Eduardo, the only change to the package is the patch to remove the hotwording feature. No other changes were introduced to the SlackBuild, the source code remained the same.
It’s weird that for this release, the Key Fix was:
“[498426] Resolved browser font magnification/scaling issue.”
It appears that this bug was not properly fixed. If you read recent comments (of some hours ago) in https://code.google.com/p/chromium/issues/detail?id=498426 you’ll notice that even the newest stable release which I am still compiling (will take until tomorrow) still has this scaling bug.
It’s weird indeed Eric. Thanks for your help.
@alienbob Please check here https://github.com/dslackw/slpkg/issues/22 is the chromium package listed twice in http://www.slackware.com/~alien/slackbuilds/PACKAGES.TXT
Hi websafe. I do not see any issue on my side. I updated that bug report with a comment.
Hey Eric,
I just wanted to drop you a note that I am having problems accessing your slackbuild repository at http://www.slackware.com/~alien/slackbuilds/.
*Edit
Sorry I guess that is all of slackware.com.
Hi Ed
Yes, there was an issue with the interaction between the slackware.com server and the Akamai contentserving front-end. We had to reboot the server in order to fix that. Should be OK now.
Yes, perfect. Thank you!
Ed
alienbob,
thank you for all your great work….
Hi again Eric,
What is the possibility of getting you to look at building VeraCrypt for Slackware-current?
Hi Ed.
I can do that.
In fact, I have just done so… veracrypt packages uploaded to the repository. Please tell me if they work.
Note that I added not just the veracrypt binary (the combined gui/console program which links against a lot of X libraries) but also a pure console binary “veracrypt-nox” which does not link against any X libraries at all and therefore can be used on headless servers that have no X installed.
Wow, thanks a lot!
I have installed it on -current/MATE and it works great! (Using GUI)
I have never experienced support like that which you give for Slackware. I am going to be making a donation to you and PV soon!
@ Alien Bob,
Many Thanks for the Veracrypt package!!!
Greatly appreciated.
I am writing to report a bug in konqueror 4.14.6 in slackware64 current. I go to my wordpress.com admin page, then to the “freshly pressed” page, and soon after the browser crashes.
Hi Darren
You should create a bug report in the KDE bug tracker instead. I do not solve application crashes.
Hello! The multilib openssl and openssl-solibs were not upgraded to correspond 64-bit versions.
Best regards and many thanks!
@Alex,
That is one of the easier things to do yourself, Download the 486, actually, I think they are labeled, 586, and follow the instructions here,
http://alien.slackbook.org/dokuwiki/doku.php?id=slackware:multilib
The information is about 1/2 way down the page.
@cwizardone, thank you for the link. In any case, alienbob already upgraded multilib.
Question about qemu
I installed Slack current from your repository, and it works very fine. thanks for that.
I would like to have virtualization env, and for that i installed latest qemu and virt-manager packages from Slackbuild. that works, but seems a more complicated and i have some problems with networking.
I read your wiki pages about qemu and Hardware virtualization with QEMU, but there’s some dead links to your packages. So can you tell me if theses howto are applicables for current Slackware.
Thanks in advance
Hi antuan –
Years ago I changed the package name in my repository from qemu to qem-kvm to indicate that this fork of qemu was the one to be used instead of qemu itself. Looking back at my archive, I had packaged qemu-1.3.0 in December 2012 because that was the first release of QEMU that incorporates all the improvements of qemu-kvm. And then apparently I never released that qemu-1.3.0 package. I cannot remember why… and I am still using qemu-kvm-1.2.0 here at home.
Essentially I am still using QEMU (read: qemu-kvm) as explained on that Wiki page. The server running the virtual machines is Slackware 13.37 however – not slackware-current.
This blog is not the best medium for troubleshooting sessions. I think it may be better to open a thread on http://www.linuxquestions.org/questions/slackware-14/
Hi Alienbob
I resolve all with Slackbuilds from repository and every thing works perfect. I just have one Question. Do you know how to do for getting updated queues file for sbopkg. Link to Gitorious don’t work at all.
And just one word to say to you that appreciate very well your work. Just Fine.
Regards
Hi antuna
The gitorious repository for queue files is no longer available, but you can use a script which is part of sbopkg (youy can find it in the contribs directory): sqg or “sbopkg queuefile generator”.
More information about how to use sqg to generate queuefiles is here: http://slackblogs.blogspot.nl/2014/01/managing-sbo-dependencies-easily.html
It was under my eyes and i don’t seen it, thanks
Hi Eric,
can you tell me if your article on qemu is of actuality. if yes, I don’t understood which packets must be installed due to dead links. I already have intalled qemu and virt-manager since slackbuilds, but I would like to try your resolution with vde. this seems more less complicated to set up. Nowadays my install works but with some problemes as for the resolutions of addresses with dnsmask and avahi which I which i cannot resolve..
Could you just to say to me quickly how you would be taken you there?
Regards
antuan, the combination of qemu and vde is what I still use, every day, when I compile a package in one of my virtual machines.
I do not use a version of qemu which needs libvirt or even avahi (how would qemu require avahi??).
Yes, concerning avahi, i quickly deleted it. I realized by consulting my syslog. Dnsmask done the job rather well, i had incorrectly configured it at the outset. Therefore everything is back in order. But I come back to my first question; if I want to go by your solution, what do I need to install? And then, why don’t you do you serve not libvirt?
HI Eric. Regarding the last packages for qbittorrent and libtorrentrasterbar (Aug 17th 2015) I have noticed that the column with FILENAMES has disappeared on qbittorrent. Would it be a bug? I have Qt5 as recommended and didn’t have this problem with the previous versions of qbittorrent. Thank you for your time.
the “Name” column disappeared from the default view indeed. Right-click the column headers and place a checkmark in the box next to Name”, that will make the “Name” column re-appear.
You might want to open a bug report to ask the developer to restore the old default.
Thank you, Eric. I will do it.
Using slackyd -d installed from SBo I noticed a possible problem with one of your current mulitlib packages.
libssl.so.0 required by:
–> Package: aaa_elflibs-compat32-14.2-x86_64-1compat32 (/usr/lib/libcurl.so.2.0.2)
–> Package: aaa_elflibs-compat32-14.2-x86_64-1compat32 (/usr/lib/libcurl.so.3.0.0)
libcrypto.so.0 required by:
–> Package: aaa_elflibs-compat32-14.2-x86_64-1compat32 (/usr/lib/libcurl.so.2.0.2)
–> Package: aaa_elflibs-compat32-14.2-x86_64-1compat32 (/usr/lib/libcurl.so.3.0.0)
What are the implications of this?
-d does “-d: Search missing shared libraries of installed packages.”, more info about slackyd can be found in the link below:
http://old.slax.org/forum.php?action=view&parentID=75894
Hi orbea
Your question has been addressed in yesterday’s -current update:
Please read the remarks about the reasons why you should not always update your aaa_elflibs if you are running slackware-current:
http://www.slackware.com/~mozes/docs/aaa_elflibs.txt
Thanks for the reply and helpful link, I have read it before and have been only updating aaa_elflibs during the bigger updates when Pat updates it.
The KWalletManager5 doesn’t show any of my wallets.
Geremia, I have several blog posts specifically about KDE5. Please discuss anything related to KDE5 in the comments section of a KDE5 post and not in the general feedback.
By the way, my kwalletmanager5 shows (and uses) my single wallet. Never used more than one.
Hi Eric. Looking at the packages i noticed -current ships with fluxbox 1.3.7 (the latest) and 14.1 ships with 1.3.5 (two releases behind). would it be possible a slackbuild of 1.3.7 for 14.1? thanks for your time and efforts.
Hi fabio.
Slackware releases only get critical security updates in their /patches directory. A newer version of fluxbox for Slackware 14.1 is something you’ll have to compile yourself.
Hello Eric, I have been reading the post “feedback” of your blog and I have learned that there are difficulties in communication when using translators of languages for the comments.I also use these translators because they don’t know English writing,read until I can but to speak and write I have no security or training good. However I see that I am using now is of good quality and I will continue to comment on using the translations but if it is difficult to understand let me know that I will try another solution
Hello Eric,
your ffmpeg package is just awesome.
The only thing it misses, IMHO, is ssl support. For example, mpv cannot play youtube videos if ffmpeg doesn’t support ssl (and, in general, all https urls). I always need to compile ffmpeg for this very reason.
Have you ever considered building it with ssl enabled? I know openssl is not feasible due to licensing problem, but what about using gnutls instead (like Arch does, for example)?
Thanks!
Hi luca
What would be needed in the SlackBuild script to add SSL support to the ffmpeg package? I have not consciously left this out, I just had not realized that this was something people could need.
Don’t worry about licensing issues, I will take openssl or gnutls, whatever is most useful.
Hello Eric,
thanks for the quick response.
it just takes to pass “–enable-gnutls” to configure to enable gnutls and “–enable-openssl” and “–enable-nonfree” for openssl.
So I think it will be forced to stay in the restricted repository if openssl is used.
I think the two are equivalent in term of functionality.
Hi luca
That will indeed be easy to add, and it also allows to compare openssl with gnutls if I add openssl only to the restricted build.
Good idea.
Hello Eric
I’ve noticed that your SlackBuild for Libreoffice 5.0.2.2 in Slackware Current includes a patch to exclude the test CppunitTest_sc_ucalc.
The failure of this test can also be replicated in calc from the GUI by following these steps –
1.
Start Calc
2.
Select a range of cells from G4 to G10
3.
Edit -> Fill -> Series
4.
Push the radio button
5.
Start Value: 5
6.
Increment: 2
7.
Expected results 5, 10, 20, 40, 80, 160, 320 in cells G4 thru G10
8.
Edit -> Undo:Fill
9.
All cells in G4 thru G10 range should now empty
I have observed the following –
A.
In LO 4.4.5 in Slackware 14.1 the above test passes
B.
In your version of LO 5.0.2.2 in Slackware Current the test fails.
C.
Using the SBO version of LO 5.0.2.2 in Slackware Current the test passes.
D.
If I compile from GIT in Slackware Current the test fails.
E.
If I compile from GIT In Lubuntu 14:04 the test passes.
Any thoughts on what might me missing in Slackware Current that might be causing this problem.
Alex
Alex, interesting issue. Indeed I could not make the compilation finish unless I disabled the unittest for sc_ucalc.
I noticed that the Fedora SPEC file for libreoffice (http://pkgs.fedoraproject.org/cgit/libreoffice.git/tree/libreoffice.spec) mentions “system mdds 0.10.3 causes a crash in sc_ucalc unit test” but Slackware does not have mdds and my build uses the mdds tarball which libreoffice downloads.
Perhaps you find the cause yourself, in that case let me know. Otherwise I will re-visit this issue when LO 5.0.3 gets released.
Eric
OK, I’ll bisect the code.
Firstly, I was wrong about stating that LO 4.4.5 in Slackware 14.1 passes it in fact fails!
What does pass is 4.3.0.3 whereas 4.4.0.3 fails.
As you can appreciate, if I do find an answer it will take some time to discover as at each bisect point I have to recompile. Can’t think of any shortcuts.
Alex
The results of the bisect, which appears to be in the right ballpark, are –
8e4dc1d760d85e09bbc3f3bbb5b8be2947db1b63 is the first bad commit
commit 8e4dc1d760d85e09bbc3f3bbb5b8be2947db1b63
Author: Noel Grandin
Date: Fri Aug 22 15:15:17 2014 +0200
create type-safe bitfield for sc insert/delete flags
The most important part of the change is in sc/inc/global.hxx
It creates a type-safe struct that prevents the accidental interaction
between regular integer types and the flags struct.
It also provides utility methods that make combining and testing the
flags type-safe.
Change-Id: Ibc5b20058b1655df913490682b679afd1297b36d
Reviewed-on: https://gerrit.libreoffice.org/11071
Reviewed-by: Eike Rathke
Tested-by: Eike Rathke
:040000 040000 f9f6a67e9269712bb1997ee6ba9fb49d21e4455e 494c585ff088b7980af64f28e556a58b2afcc35b M sc
Eric
I posted a message reporting my findings to the Libreoffice developers list on 24/10/2015 regarding the patch “create type-safe bitfield for sc insert/delete flags” not working on Slackware but was working on Lubuntu.
Low and behold a complete rewrite of the patch appeared 2 days later 26/10/2015 called “convert InsertDeleteFlags to use o3tl::typed_flags”.
I’ve compiled and tested in master. Now all tests pass in Slackware. Can’t find a bug report for it so don’t know what version of LO will end up with the patch.
Alex
Hi Alex,
That was a fast fix!
I will check the next release to see if this patch has been applied or else I can apply it myself.
LO 5.0.3 should be released very soon according to https://wiki.documentfoundation.org/ReleasePlan/5.0#5.0.3_release
Hi Eric
Not quite out of the woods yet.
Tracked down 2 regressions in LO 5.0.2 still got two more to go before I can move off of LO 4.4.5.
The one here was quite trivial, the other three are show stoppers for me.
One I’ve raised a bug for and that’s been fixed – 93618.
One of the unreported bugs locks up my virtual machine in virtual box. Have to go to tty1 to kill it.
Every time they bring out a major release my base application stops working. The bugs get fixed by about x.0.4, If I report them that is.
Alex
Hello Eric.
Yesturday I had odd stoper when tryed to update slackware 14.1 x86_64. After both apr*txz packages in xterm shows somthing was gine wrong and there is no /bin/sh. After that system just become dead. I reinstalled compleatly system and trayed agayn update and got same result. 🙁 So for now I instaled slackware64-current.
I wonders is this issue just striked me?
P.S but anyway thanks for grate and stable distro. 🙂
Hi Renars
What was the Slackware mirror where you got your slackware64-14.1 packages from? This issue should not occur with the packages in the slackware64-14.1 tree unless the package tree got corrupted.
Unless you mistakenly used a 32-bit package repository when installing these updates, instead of the required 64-bit repository.
Shame on me!!! 🙁 I really mistakenly used 32bit packages. Well than tahnx for Your time and effort to answer for so fullish mistakes 🙂
KDE 5 is running fine here, thank you for this much of work!
Hello, I just finished trying to build your SlackBuild for ffmpeg. It seems like it’s got some problems. I’ll paste in the last part of the terminal output so perhaps that will be of some help. I think this is the first one of your SlackBuilds that hasn’t compiled for me. Hope this helps.
Oh, btw, I downloaded everything in the build directory and kept the directory structure exactly the same as you have it on the site.
make[1]: Leaving directory `/tmp/build/tmp-ffmpeg/zvbi-0.2.35/po’
Making install in doc
make[1]: Entering directory `/tmp/build/tmp-ffmpeg/zvbi-0.2.35/doc’
make[2]: Entering directory `/tmp/build/tmp-ffmpeg/zvbi-0.2.35/doc’
make[2]: Nothing to be done for `install-exec-am’.
make[2]: Nothing to be done for `install-data-am’.
make[2]: Leaving directory `/tmp/build/tmp-ffmpeg/zvbi-0.2.35/doc’
make[1]: Leaving directory `/tmp/build/tmp-ffmpeg/zvbi-0.2.35/doc’
make[1]: Entering directory `/tmp/build/tmp-ffmpeg/zvbi-0.2.35′
make[2]: Entering directory `/tmp/build/tmp-ffmpeg/zvbi-0.2.35′
make[2]: Nothing to be done for `install-exec-am’.
test -z “/tmp/build/tmp-ffmpeg/ffmpegdeps/usr/lib64/pkgconfig” || mkdir -p — “/tmp/build/tmp-ffmpeg/ffmpegdeps/usr/lib64/pkgconfig”
/usr/bin/ginstall -c -m 644 ‘zvbi-0.2.pc’ ‘/tmp/build/tmp-ffmpeg/ffmpegdeps/usr/lib64/pkgconfig/zvbi-0.2.pc’
make[2]: Leaving directory `/tmp/build/tmp-ffmpeg/zvbi-0.2.35′
make[1]: Leaving directory `/tmp/build/tmp-ffmpeg/zvbi-0.2.35′
/tmp/build/tmp-ffmpeg
**
** ffmpeg …
**
ERROR: libass not found using pkg-config
If you think configure made a mistake, make sure you are using the latest
version from Git. If the latest version fails, report the problem to the
ffmpeg-user@ffmpeg.org mailing list or IRC #ffmpeg on irc.freenode.net.
Include the log file “config.log” produced by configure as this will help
solve the problem.
Makefile:2: config.mak: No such file or directory
Makefile:62: /common.mak: No such file or directory
Makefile:104: /libavutil/Makefile: No such file or directory
Makefile:104: /library.mak: No such file or directory
Makefile:106: /doc/Makefile: No such file or directory
Makefile:189: /tests/Makefile: No such file or directory
make: *** No rule to make target `/tests/Makefile’. Stop.
Makefile:2: config.mak: No such file or directory
Makefile:62: /common.mak: No such file or directory
Makefile:104: /libavutil/Makefile: No such file or directory
Makefile:104: /library.mak: No such file or directory
Makefile:106: /doc/Makefile: No such file or directory
Makefile:189: /tests/Makefile: No such file or directory
make: *** No rule to make target `/tests/Makefile’. Stop.
** Adding ffmpeg presets:
** Fixing up pkgconfig files:
ls: cannot access /tmp/build/package-ffmpeg/usr/lib64/pkgconfig/*.pc: No such file or directory
cp: cannot stat ‘doc/TODO’: No such file or directory
cp: cannot stat ‘doc/*.html’: No such file or directory
./ffmpeg.SlackBuild FAILED at line 1765!
bash-4.2$
The only relevant message shown in that bit of text is “ERROR: libass not found using pkg-config”. This means something was not well, a lot earlier in the compilation process. Try finding out what happened when libass was being compiled.
Build ffmpeg from the Github repo; it builds flawlessly, and has the advantage that it is compatible with the GH repo for vlc, too, and it was the only way I could get the latter to build successfuly.
Alex Kempshall,
In LibreOffice 5.0.3 the patch http://cgit.freedesktop.org/libreoffice/core/commit/?id=eaa34b4e29478770c4707d56aa102491bc8e058a still has not been applied, and I am unable to apply it cleany against the 5.0.3.2 tarball. Subsequently the compilation fails because of the partly failed patch.
Unless you are able to create a rebased patch for me, I am afraid I have to build LO 5.0.3 without this fix.
Actually Alex, I will try to come up with the patch myself. Most of the failed stuff is a simple s/IDF_/InsertDeleteFlags::/g
Hi
Before raising it with you I posted a message to the developers list. Didn’t get anything useful back.
Then bisected the code got it down to a specific patch. Raised it again on the developers’ list and another patch was raised, a couple of days later, against master which fixed the problem.
I’ve never managed to find a bug report so don’t know which branch/release the patch was aimed at.
I’ve asked on the Developer’s list.
Alex
I have a patch now, and I am compiling.
Either I missed parts of the patch in git or something else is not compatible. This is the error I get:
414461 :/tmp/build/tmp-libreoffice/libreoffice-5.0.3.2/sc/inc/global.hxx:152:11:
error: 'fromInt' is not a member of 'InsertDeleteFlags'
414464 :/tmp/build/tmp-libreoffice/libreoffice-5.0.3.2/sc/inc/global.hxx:152:38:
error: 'lhs' was not declared in this scope
414467 :/tmp/build/tmp-libreoffice/libreoffice-5.0.3.2/sc/inc/global.hxx:152:50:
error: 'rhs' was not declared in this scope
The patch I used is here: https://dl.dropboxusercontent.com/u/2329942/libreoffice5.failing_unittest.patch
Hi Eric
the response from the developers is
“as it was pushed to master only recently, it will only appear in the next release spawned off master, i.e., LO 5.1 ”
So I suppose we should leave it to spawn out in the way the developers intended.
Maybe if I’d raised a bug report immediately I encountered the problem the developer may have back ported.
Alex
Dear Alien Bob,
Would you mind update Qt on stable repo?
It’s a long time that Qt is updated to 5.5 on slackbuilds.
Also Qt has been updated to 5.5.1 about 30 days ago, If it’s possible, would you mind udpate to this version?
Thanks a lot,
Hi Eric! It will be nice if you will also make a persistent Slackware Live edition.
People like the idea of having a stable and powerfull live operating system stored on an usb stick.
Daniel, did you not read my article about my Live edition? Probably not, since you are posting this in the generic Feedback section.
Hi! After some years of being away from Slackware (I got seduced away by Debian and the lazy ease of apt-get update) I find myself not loving the news of systemd and looking to return to Slack.
BUT! The question is how long can systemd be held at bay? Estimates? My current laptop is nearly dead and I’d hate to buy a new one and find myself hamstrung by systemd issues. If I must be punished for using a computer I’ll give in and go Apple. But if you figure systemd can be held off 3-5 years I can hold off on such decisions.
Systemd will never be a requirement to run a computer. Slackware current recently replaced its old udev with eudev, so its seems like slackware will not have systemd in the near future if ever.
Chris, honestly.
I can not look 5 years into the future of Slackware. Make up your mind about what you _want_ to use and then stick with that decision.
Fair enough, then Mr Hamleers! I over-reached with that phrasing. I was more after seeing what someone like yourself with more insight into the development side of things felt about the potential down the road. I am comforted seeing the work being done by Slackware, XFCE, Gentoo and (I think) Funtoo.
@orbea – Originally I wasn’t particularly afraid of systemd – I figured it was one init system amongst many and linux tends to allow multiple choices. But it seems like (looking at KDE and GNOME) that some of the DEs are hooking too much into systemd and every other distro has jumped on the bandwagon shouting “Newer and shinier!”
And sytemd has long since shot past being a simple init system and taking over other functionality. So it started to feel a bit inevitable.
Hey alienbob, maybe you can consider adding lzo and libaio to your multilib repo? I was investigating /usr/lib with ldd and found that libcairo-script-interpreter.so.2.11400.4 owned by cairo and libmysqld.so.18 owned by mariadb link against liblzo2.so.2 and and libaio.so.1 respectively owned by lzo and libaio. Using convertpkg-compat32 on both and then installing them fixed the issue as far as ldd can see.
Hi orbea,
Sure. I added the two to my script, and I wanted to do an update before the weekend anyway. So hopefully these two new compat32 packages will be available for download by the end of the day.
So I was investigating with ldd more, then I found this old script to find missing libraries.
https://bbs.archlinux.org/viewtopic.php?id=6460
Which I then touched up slightly for slackware.
http://pastebin.com/kWYVT0cu
So in short, these compat32 packages in your repo have missing libraries which you might want to consider fixing.
alsa-lib
cups-filters
flac
gnome-keyring
gnutls
gst-plugins-base
gst-plugins-base0
gst-plugins-good
gst-plugins-good0
libgphoto
libxml2
libxslt
mariadb
qt
util-linux
So I then used convertpkg-compat32 to create packages containing all the missing libraries listed and then repeated a few times to make sure I had them all. In the end installing these compat32 packages resolve all missing multilib libraries.
GConf
aalib
at-spi2-atk
at-spi2-core
cdparanoia
db48
gc
gcr
gd
glib
gtk+
gtk+3
guile
judy
libcaca
libiodbc
libsoup
libtheora
libunistring
libvisual
libvpx
poppler
python
qpdf
taglib
tcl
tk
wavpack
xmms
And here are more notes for this which might help.
http://pastebin.com/skq3urqp
Hi orbea,
The compat32 packages are not meant to be satisfying all possible library dependencies. They are meant to get your proprietary 32bit programs running, and those programs will not tap into all the functionality. Whenever someone asks to add a package to the list because some 32bit program refuses to run otherwise, I do that. By following the path you entered above, you will eventually end up with a huge list of packages to install, to a point where you might consider just installing full 32bit Slackware in a chroot and be done with it.
Fair enough, I thought I’d mention it regardless. 🙂
Hello alienbob,
I think this article is about slackware.
I think the guys are very happy with slackware in Enterprise, and give some advises, good ones.
Slack won\’t — and shouldn\’t — play IT\’s game
http://www.infoworld.com/article/3012118/saas/slack-wont-and-shouldnt-play-it-game.html
I think he is just carrying about business and enterprise, as Slackware is growing in a new land (that is awesome).
I think Pat and you will be happy with this article.
If is not about Slackware, feel at it was.
Only knows that, everybody root for Slackware, and I hope back to use it soon.
Phantom, you should read better. That article is NOT about Slackware Linux.
Hi, Eric. I wrote a few months ago about problems I’d been having running Chromium after version … uhm, it’s been so long, I forgot. Anyway, after installing your latest Chromium (47.0.2526.73), here I am, writing you in Chromium! No more “Aw, snap!”, and no more having to resort to “chromium –disable-seccomp-filter-sandbox &”. As a Bugs Bunny character once siad, “I don’t know what yuz done, but I know yuz done it.” 🙂 Thanks, man!
Hi alienbob,
Have you ever thought about supplying an extra multilib repo for packages on SBo that might be needed as a multilib version? One example is pcsx2 (Playstation 2 emulator) requires multilib versions of wxGTK/wxPython, nvidia-cg-toolkit, portaudio and soundtouch which I suspect is the main reason that its not on SBo yet.
I could just make a 32-bit chroot to either run pcsx2 in or make the required multlib packages myself, but you might be in a better position to do this for a greater number of people?
Hi orbea
Thanks but no… I am busy enough with just my own stuff. People will have to build their own “compat32” versions of SBo packages if they need that – it is not difficult.
I downloaded the 64bit slackware-current ISO on 12/13, and there has been a big change for me: I can’t get it to boot, either on my laptop or as virtual machine instances, unless I add the line “maxcpus=1” to the LILO append options. I haven’t been able to find any documentation as to why this is, whether it’s a worry, or if it is what to do about it. Once I do that all else seems to be just fine.
Hi Capouch; I never heard that before. There has not been a change in how the ISO is being generated, but it is of course using a newer kernel to boot than in the past. Perhaps this is a kernel issue? What kind of errors did you get when booting from the ISO without the “maxcpus=1” parameter?
Also curious about how the installed system is behaving. Is the “maxcpus=1” parameter not required there?
The ISO boots a “huge” kernel, is this the same kernel you are using on the installed computer, or did you switch to a generic kernel plus initrd?
It is the default kernel to boot (huge) and whatever kernel is installed in a full install, I suspect the same one.
I can send a screenshot via URL of the screen as it boots? Although I’ll have to take a picture of it–it is something about detecting CPUS then a new line then #1 some spaces #2 and a freeze. I can see no ill effects of the maxcpus. We refresh our current pretty often. The late October one was just fine, but the mid-December one, as I said, requires that append for both the original install and the same ISO installed under VirtualBox.
Hi again alienbob,
In short: Any chance you could add libnotify and xz to your compat32 repo?
Explanation: Pcsx2’s last stable tarball is ~1,500 commits behind and does not work well, so I made a git version of the slackbuild which does work. However it requires two Slackware compat32 packages you do not already include in your repo. Pcsx2 requires xz and wxGTK3 builds against libnotify which will result in pcsx2 failing to compile on a multilib environment without it.
They are both easy to make by the user, so I just included them in the slackbuild’s README. The slackbuild is available in my small notabug SlackBuilds-git repo in the following link.
https://notabug.org/orbea/SlackBuilds-git/src/master/pcsx2
Hi Eric,
I just wanted to pass on that your vlc-2.2.1 build script failed to complete on Slackware64-current until I added the following patch.
— a/vlc.SlackBuild 2016-01-03 21:16:39.782066543 +1100
+++ b/vlc.SlackBuild 2016-01-03 21:17:02.035064258 +1100
@@ -2924,7 +2924,7 @@
{
echo -e “**\n** dvbpsi …\n**”
cd $TMP/tmp-$PRGNAM/libdvbpsi-${DVBPSI}
-CFLAGS=”$SLKCFLAGS” \
+CFLAGS=”$SLKCFLAGS -Wno-error=bool-compare” \
CXXFLAGS=”$SLKCFLAGS” \
LDFLAGS=”$SLKLDFLAGS” \
./configure \
As always, my grateful thanks for all you do for the Slackware community.
Hi Eric,
this is a question on Slack 14.1 Installation with UEFI/secure boot disabled. Is there a link to understand why write fails?
cgdisk 0.8.7
Disk drive /dev/sda
Size …
Partition # Size PartitionType Partition
1007.0KiB free space
1 400MiB Windows RE Basic Data Partition
2 260MiB EFI System EFI System Partition
3 240MiB Microsoft Reserved Microsoft Reserved Partition
4 345.4GiB Microsoft Basic Data Basic Data Partition
7 168.3GiB Linux filesystem Slack14.1
9 736.0KiB Linux swap
8 168.2GiB Linux filesystem (I intend to also use another Linux distribution here)
5 450.0MiB Windows RE
6 15.3GiB Microsoft Basic Data Basic Data Partition
871.5KiB free space
After setting up partition 7, verify was ok and then write says: Problem saving data! Your partition table may be damaged!
Hi Svar
My blog is no extension of LQ.org. You already posted your question as http://www.linuxquestions.org/questions/slackware-installation-40/gpt-issue-on-a-14-1-slack-install-on-laptop-4175563062/ and it has answers from fellow Slackers. That is how it works – and if no further answers are posted that’s probably because no one knows.
Expect nothing here because I am not running a free for all troubleshooting shop. Sorry.
Hi allend,
I will probably need that when I compile vlc 2.2.2… whenever *that* comes out.
Eric,
thanks both for your answer and your huge slack work. I’ll post in LQ if I figure it out.
Svar, I may have sounded a bit harsh, which was not my intention. Hope you get your answer on LQ.
Dear Eric,
Looking to introduce/unlock children at the community level to computers and what not, etc. of course running Slackware.
This also applies to my children who are yet to ever touch a computer.
I kindly request if your “Parental control on the Linux desktop” from 2008, would be still how you would go about controlling things as such.
http://alien.slackbook.org/dokuwiki/doku.php?id=slackware:parentalcontrol
warm regards
Hi _metic
My son has grown up and we are no longer in need of parental controls here, but if I had to do it again, I would start with my old instructions and see how far I would get. I hope that the old Wiki article is still 100% valid but I cannot give you a guarantee.
What I learnt over the years is that installing parental controls is the start, not the end.
The more important part of acquainting your kids with the Internet and keeping them away from its dangers is to keep the dialogue with your kids open. Talk with them about what they are doing with their computer, be genuinely interested and hope that your interest will keep them informing you of what they do.
Once they start getting the feeling that you are tieing them down they will start exploring without telling you. And smart kids will eventually find their way around the parental controls.
Still, these parental controls helped a lot.
thank you for the response.
Hi Eric, in case you don’t already know, Copy.com will be discontinued on May 1:
https://techlib.barracuda.com/display/COPY/Copy+End-of-Life
Bad news.
Regards,
Eduardo
That’s too bad indeed Eduardo. I had not yet seen this EOL message.
http://taper.alienbase.nl is down 🙁
Hi alienbob.. i love your contribution to slackware comunity.. but i had a problem installing multilib on my slack64 with your tutorial.. once i completed the installation all of the 32bit programs worked.. but when i restarted my Xorg i couldn’t get back in.. no matter what config i used everytime i run startx all it did was post a list of libraries and thats it.. so i had to reinstall the system all over because had no idea what the problem was. do you know what did i do wrong? using 14.1 slack with kde installed from the latest dvd
Anze – you are running Slackware64 14.1 with the KDE 4.10.5 from that same Slackware?
And after you installed the multilib packages, your X session would no longer start? The most common reason for that is that you installed glibc for the wrong Slackware version. From what exact URL did you download these multilib packages? And again, is it Slackware 14.1 or -current or a mix of the two that you have installed?
hi bob.. i was following your tutorial http://alien.slackbook.org/dokuwiki/doku.php?id=slackware:multilib
yes all from same slackware. and was downloading slackware 14.1 packages.
i ran a command:
# SLACKVER=14.1
# mkdir multilib
# cd multilib
# lftp -c “open http://taper.alienbase.nl/mirrors/people/alien/multilib/ ; mirror -c -e ${SLACKVER}”
# cd ${SLACKVER}
and all of the files were downloaded for 14.1 version of slack. i followed exactly your tutorial and xorg failed. now i’m afraid to do it again
Anze, does not sound wrong what you did. But again, what exact version of Slackware are you running? What medium did you use to install? The official Slackware 14.1 DVD ISO image?
hi bob. i am running slack 64bit 14.1 installed from official kde dvd downloaded from slackware website.
Hello.
Your last change log is time stamped about 12 or 13 hours ago, butt none of the packages have hit the mirrors I’ve looked at.
Just FYI.
You’re right. Apparently I did not add the packages to the repository. Attribute that to the spell of flu I am under.
Understand. I’m in my 5th week of the flu. The doctor said there is a strain going around that has hit the elderly (moi) very hard and there isn’t much that can be done about, but tough it out. The worst is over, but the cough won’t go away….. Getting real tired of it.
🙂
Sorry to hear that Eric.
Hope you get better really soon.
And, thanks for all the hard work you do!
Hi
Icedta 2.6.4 is released; however the repository version of icedtea is stopped at 1.6.1 which is out of date and is not designed to work with opendjdk 7. I would be grateful if you update the iceadtea package.
Thank you for all your effort.
Mehran you should check your facts better before you post here. I do not have an “icedtea-1.6.1” package at all.
I have a package called “icedtea-web” at version 1.6.1 and that contains the browser plugin for Java applets. There was a newer release last week (1.6.2) which is on my TODO but not with a high priority.
I do not have a separate “icedtea” package, that would be useless. I have an “openjdk” package which is currently at version 7u95_b00 and which I compiled using icedtea 2.6.4.
You’re right. I’m sorry for this mistake.
Hi bob. i think i found a solution to my problem. Faulty fglrx driver. I tried one more time to make my 64bit multilib and xorg failed yet again.. i had enough of it and installed 32bit slack… i updated the system with slackpkg and when it was done i rebooted the system. when it booted back into slack i started Xorg and bam the same problem allover again.. than just for kicks i rerun the amd video card drivers installation script and xorg started working again..so now everytime a mess with lib files i have to rerun script to make xorg working again.. so thanks for your time trying to find a solution for me.
Hi,
Just wondering if you had my e-mail to alien at slackware dot com about a possible mirror. I haven’t any error but don’t know how often you check it. (it’s from another e-mail address)
Hi Tonus, I think I received your email, and another too with a mirror proposal. Currently busy due to real-life situation but I _will_ respond. Thanks!
I wonder if you know about http://ring.cx I just discovered it recently. It is a secure skype replacement.
With a qt5 dependency how hard would it be to add to slackware?
Adding to Slackware core would be out of the question, looking at the plethora of dependencies that need to be fulfilled first. But it should be feasible to build this and its deps and put it in a 3rd-party repository.
Interesting, because open source voice/video capable peer-to-peer applications are rare in Linux world.
@alienbob, I am looking into nix pkg manager to see if that can be used but so far it has given me a headache. And yes ring.cx is a rare thing and is why I watched the fosdem talk.
Eric,
Four years ago I convinced my school to let me teach the 8th graders about computers by doing a 3 week course — 2hours/morning. We had no computers, and no budget for software, so I solicited laptop donations from the community. I was a former IT professional who used Unix (some) at work. I had never used Linux, but I knew I could get it for free (which was perfect for my budget). I wiped the laptops and wanted to laod Slackware 13.37 on them. I received everything from a 1999 IBM ThinkPad to a ASUS EEE netbook. Back then, I was totally stuck how to get slackware installed since I had no disks, had never heard of an iso image, and didn’t have a Linux installation on any computer. Your http://alien.slackbook.org/blog/welcome-windows-user/ link helped me quite a bit. I was able to create a USB stick that loaded my motley collection of PCs. (OK, I had to figure how to burn disks for the Thinkpad). The course was a great success. I taught the kids the basics of programming using only bash and a litle python. The kids loved it, and the course continues even though I am no longer teaching there.
This year I am teaching the course in a new city to more 8th graders. Familiar with the process, a couple of weeks ago, I started collecting old laptops to load them with Slackware. Two days ago my USB stick died. (It was 3 years old.) Reluctantly, I sought to re-create the bootable install stick. I had hot-wired the first one with a universal linux installer and your advice. I was never quite sure how I got it to work, so I was very unhappy about re-creating it. However, I now have several laptops with Linux on them, so I could create bootable USB from that, right?
This led me to your post http://alien.slackbook.org/blog/installing-slackware-using-usb-thumb-drive/. I used the information there to create new stick. Note, I got hung up along the way because I specified -o /dev/sdb1 in the usbimg2disk.sh command, but I found my mistake in your response to a comment on May 29th, 2010. I also had to install the rsync package as you suggested on October 7, 2010.
So you never even had to talk to me to help me. I just thought you should know that you, like Richard Stallman and Linux Torvalds, made my work possible. So far, you’ve helped bring progamming to more than 85 middle schoolers that you never even knew about (and who knows how many more in the future).
So, thank you.
Who knows how many people you are helping every day without even knowing about it?
Sincerely,
Randy Evans
PS I don’t actaully have a question or problem. You’ve answered all of them (so far).
Hi Randy
Nice of you to drop a note and tell us about your endeavors. Nice story! I think teaching kids the art of computer programming at a young age will create brilliant people when they grow up. It’s like with the Raspberry Pi which was conceived for exactly the same reason: create an affordable platform where kids can be introduced to computer programming in a playful manner. the world will need these kids sooner or later.
The world is getting dumber by the day and tracks like yours are fighting that awful trend to kill creativity by turning young people into consumers instead of thinkers.
Did you write your course program from scratch or could you use existing coursework?
I made it up/make it up as I go a long. It’s very discovery based which, for me, gets the best results. It’s all command line. On day one, for example, they power on the laptop, I give them a username and paswword and ask them to make the computer do something. Every year, I amazed at what they figure out from that. New info is traded up and down the rows, and a few reboots become necessary. Eventually, I have to give them some direction, but once you get hooked on the notion that there are secrets in there that you can master, the computer becomes an interactive trial and error machine.
As a teacher, I can’t resist imposing some structure and having some markers along the way, but the more open-ended it is the more the kids are engaged and the more they learn.
I’ve been thinking that the Rasperry Pi would be the way to go long term. I’d love for the students to see the guts a bit more to help disple the magic box myth, but a laptops is self-contained, less breakable, and so far free. Kids can take them to and from school quite easily. Pi would be hard to handle with all the peripherals.
Thanks again for your help.
hi eric, do u noticed the last chromium-dev-pepperflash-plugin-20.0.0.248-x86_64-1alien.txz causes noisy and distorted audio on chromium flash video players?
Hi inman; to be honest, I do not use Flash myself so I can not comment on distorted audio. I wonder, are you running it on Slackware 14.1 or on -current? Or alternatively phrased: is your computer using ALSA, or PulseAudio, or perhaps even JackD?
Openshot is one of the best free video editors out. unfortunately, the one provided at SBo has a quite extensive list of dependencies and optional dependencies. I really would like, if possible, to see it built in a single txz package. And thanks for the new liveslack, using it as pocket os with full disk encryption.
hi, tanx for the response here …
I’m using the SC with pulseaudio, so probably the chromium-dev uses alsa playback plugin. The html5 players (eg youtube) are working perfectly.
The noisy and distorted audio ( like audio glitch, sounds like corrupted mp3) is only there if a flash video or audio is being played using chromium-dev. With seamonkey the problem isn’t there at all. So probably it has not to do with pulseaudio but the flash plugins.
The flashplayer-plugin-11.2.202.569-x86_64-1alien and seamonkey doing well, while the chromium-dev-pepperflash-plugin-20.0.0.248-x86_64-1alien and chromium-dev-49.0.2593.0-x86_64-1alien have some problem.
I have some guesses about f_s, but i couldn’t figure out how to check or configure that.
If u do not use the flash-player at all, how to you mange to play the video players of those creepy old fashioned website? … is there anything out there that im not aware of!
thank u as always, for all you do for slackware and its community.
bests
inman, some people do not visit creepy websites with flash video.
And if I visit pages with flash-based advertisements, they usually do not come with audio so I would not notice that something is wrong with the audio…
Even if it is bbc.co.uk? i mean where can you find unbiased news with html5 player? 🙁 😀
For unbiased news I read a news paper. Real paper.
Thanks for all of your hard work on Slackware. Are we going to see LibreSSL enter the main repository, if not replace OpenSSL outright? Also, will the Lumina desktop environment ever make its way into the repositories? Thanks, Ryan.
Hi ryan; no and no…
@alienbob
THANK YOU…to you and all those that make slackware what it is !
Keep up the great work !!
Hi blizzack – thanks for the kind words. Glad you like Slackware, be sure to advocate it among your friends 🙂
Hello Alien .. you are a legend;)
I would like to ask you how can I make a non-root user login on KDE 5_16.02
Thanks Francesco
darkstaring , that is not limited to KDE, it is a Slackware feature to create a non-root login right after installation of the OS.
Read http://docs.slackware.com/start?id=slackware:beginners_guide#create_a_user_account
Once you have a non-root user it should be able to login to the SDDM session manager, or if you are in runlevel 3 (console login) you first run “xwmconfig” (once) to select xinitrc.plasma as your session type. Then run “startx” everytime you want to start an X session.
Hi Eric
Thought you would be interested to know that a member of the Puppy Linux community has found a way to obtain 32-bit pepperflash updates from Google Chrome. How long this will continue to work is of course unknown but its has worked for 21.0.0.182.
The thread describing the process is: http://murga-linux.com/puppy/viewtopic.php?p=894322#894322
BTW – your update process seems to have deleted the 32-bit pepperflash in your repository.
Thanks for all you do to provide software for the Linux community.
Cheers
peebee
peebee – the 32-bit PepperFlash in my repository was vulnerable to attack. That is why I removed it.
Interesting thought that the older 32-bit Chrome still downloads an uptodate PepperFlash library from somewhere.
Hey Eric, I put up a mirror of slackware live last fall but it stopped syncing last month.
I’m getting: "@ERROR: access denied to mirrors from UNKNOWN " when I try to run sync manually.
Do you need to whitelist my IP (208.91.197.168) or do you have enough mirrors and I should take it down?
(this isn’t really a comment… I just didn’t know how else to reach you)
Cheers,
Paul
Hi paul, this was discussed in several recent blog articles. The server’s rsync has been made unavailable because it was not able to cope with the rise in download requests and my hoster started complaining about the bandwidth cost.
I have brought a new server online in its stead and you should start mirroring from that one: rsync://bear.alienbase.nl/mirrors/
I am engaged in some back-and-forth about Slackware’s "inferiority" over the issue that is pitching this error on a build:
** snip **
/usr/bin/ld: this linker was not configured to use sysroots
**
The folks on the mailing list say, "The linkers on all the other distros will do this. Your problem is Slackware. I hope you can comment–I’m not familiar with the concepts that are being discussed. Thanks.
brianc – without any context I can not give you a meaningful answer. I never saw the error, I also do not know what triggers it.
BUt… if you are in a discussion with some developer morons who use phrases like "inferiority" there is little chance that I will have a second look at what their crappy code does.
FYI, please do not post complete build logs in this blog, I will delete your posts. Use a pastebin if you want to share large chunks of text.
alienbob, is it okay to use bear.alienbase.nl for getting multilib?
It’s OK to use bear.alienbase.nl for downloading EVERYTHING it carries. This server has taken over all the repositories from taper.alienbase.nl (except for the historical Slackware releases) and this weekend I also plan on moving docs.slackware.com from taper to bear.
That did it. Thanks.
http://mirrors.edgeorbis.com/distros/slackware-live/
I solved my issue by rebuilding ld from source, version 2.26.20160125. The older version was 2.25.1. I built with vanilla configure, so I wonder if the default has changed to include the sysroots feature? In any event, it looks like it will work itself out in later releases of Slack, and I am now good to go. Thanks.
i know there’s a lot of things to do in your real life sr. That said, there is a new version of veracrypt out. no serious bugs for linux found, but there are speed improvements according to the release notes. So, here’s my request for a slackbuild of that.
Thx for the pkgs.
After painting myself in and out of dual version Qt hell a couple times I discovered your stuff.
I had heard of your pkg’s but ‘old hunt for dependencies and compile’ habits die hard.
I got so far into the corner that I reblasted 14.1 and re-slackpkg’d myself back to ‘current’.
(tolip hangs head in shame for using default microsoft troubleshooting procedure)
I noticed a difference between your Qt5 and the .run file version that throws everything in /opt.
Symptom:
When starting a Qt5 app it fails with,
This application failed to start because it could not find or load the Qt platform plugin "xcb"
Investigate:
root#> ldd /usr/lib64/qt5/plugins/platforms/libqxcb.so
It listed 4 file not found issues involving
libxkbcommon
Solution:
git and make libxkbcommon
I just copied them to /usr/lib64 manually once built.
root#> ldconfig
Summary:
I have no idea if this is ‘as expected’ or what, but it’s what I did to make it go. I’m moving on but I figured I document what I did, here, just in case it is relevant to your endeavor.
Now I’m off to play with the must have Qt5 app that started this snowball rolling 🙂
My qt5 package picked up an unexpected dependency on my libxkbcommon package… because I guess I forgot to de-install from my build box.
The libxkbcommon package is part of my KDE "ktown" repository: http://bear.alienbase.nl/mirrors/alien-kde/current/testing/x86_64/deps/libxkbcommon-0.5.0-x86_64-1alien.txz and http://bear.alienbase.nl/mirrors/alien-kde/current/testing/x86/deps/libxkbcommon-0.5.0-i486-1alien.txz
Hi,
Am not new to Slackware as it was my first linux back around 1998 or so. Am running current64. Took the time today to follow your directions to change to KDE 5. For some reason when I used:
# upgradepkg –reinstall –install-new x86_64/ . . .
it ended up bellyaching to me that it can’t find any installed files to upgrade so it skipped them. Not all of them but lots so I just waited until everything was done and then went back to just installpkg all. It worked for me.
Also, just found that dolphin has very few icons in the icon view. Don’t know what to do about that.
Plasma is breathtaking. Thanks for your hard work
Nitesky, "upgradepkg –reinstall –install-new" will install all packages that were not installed before (the "–install-new" parameter takes care of that) so additionally running "installpkg" should not be necessary. Check your /varlog/packages directory for duplicate packages, pretty sure that all of the new Plasma packages will show up.
If you have zero duplicate packages then the following command should have zero output:
ls /var/log/packages/ |rev |cut -d/ -f1 |cut -d- -f4- |rev |sort |uniq -dYa, got zero when using that command.
"upgradepkg –reinstall –install-new" for some reason doesn’t work for me so I just cover after by installpkg with everything. All new packages just install and all previously installed packages install over top again. Just a way to get the result I need.
Found konqueror lacks a menu entry. Had to make one
The command
ls /var/log/packages/ |rev |cut -d/ -f1 |cut -d- -f4- |rev |sort |uniq -d
produced zero results.
I don’t know why the command
upgradepkg –reinstall –install-new
does not install new programs for me but that is the reality. Been like that for a long time now and through a couple of clean full installs. No biggie. My way works for me
If there is a better way to report problems just let me know.
Can’t find a launcher for Dolphin
This is a KDE problem so I will ask it here. Since the last upgrade to -current I have a problem that KDE thinks that my time zone is "Troll/Antarctica". My local time zone in Slackware is set to Asia/Chongqing, which is correct (I’m in Beijing). XFCE has the correct time and time zone is "local", but KDE thinks local is Troll, Antarctica, and sets my local time 8 hours ahead, which does interesting things to my email.
I had this time zone problem before but it was easily corrected, but now KDE keeps switching me back to Troll. Humourous it is but now I am ready to have it back to droll normality.
E. Wayne Johnson, I have no idea. You should ask on the LQ forum perhaps, or file a bug on bugs.kde.org.
I seem to have fixed it by a combination of running hwclock –show which said that I was on UTC time although I had just a little while before run pkgtool to set the time. I then ran timeconfig directly from root, then hwclock –systohc –localtime which fixed it. I did timeconfig through pkgtool previously but didnt fix it for me.
My question concerns whether it is possible to have a local keyfile for a luks encrypted disk. I have an X200 with libreboot. The advantage of libreboot is that grub is my bootloader and permits full disk encryption including the /boot directory. I was able to follow your guide and have everything working great. The only tiny quirk is I have to enter my luks password twice. Once when grub first boots, and a second time during the kernel boot process. I followed a modified version of the instructions here https://libreboot.org/docs/gnulinux/encrypted_parabola.html to get grub booting. Towards the bottom of the page there is a Bonus section. It shows how to use mkinitcpio to permit a luks keyfile to sit on the local drive and be used to decrypt the drive during boot with the help of grub. The issue of course is Slackware has mkinitrd; which from what I understand requires a separate drive that’s a FAT partition for this to work. Is there a way to place the luks keyfile locally and somehow specify it with mkinitrd; or a way around this?
Boris, anything is possible with the init script in the initrd image. You should be able to add support for a LUKS keyfile on any filesystem that can be read by the init script at boot. If the /boot partition is already unlocked at that time you should be able to mount it and read the keyfile.
Eric, how would I change this line to make it work?
mkinitrd -c -k 4.4.8-smp -m ext4 -f ext4 -r /dev/cryptvg/root -C /dev/sdx2 -L -K LABEL=TRAVELSTICK:/keys/alien.luks
Even the instructions of mkinitrd say use a FAT filesystem
You have to work this out yourself.
The mkinitrd command is not what you need to change.
It is the package source. You have to edit the “init” script which is part of the mkinitrd source, add support for a LUKS keyfile on another filesystem than a (v)FAT USB stick, make sure the init script mounts that filesystem and uses your keyfile; and then you have to rebuild the mkinitrd package and install it to your computer,
An exercise I will leave to you.
don’t know if I’m misunderstanding, but all the partitions are encrypted including /boot. I can’t mount any devices unless I provide a password or keyfile first. So don’t I have to add the created keyfile into the initrd package? ; because I wouldn’t be able to access it locally.
I do not know how the full disk encryption (including /boot) works with GRUB, but you need to enter your LUKS password to let GRUB access the kernel and initrd on the /boot partition, right? Doesn’t that mean that from that point onwards, /boot is unlocked? Or is /boot only accessible to GRUB during initial boot? If that is the case then indeed you have to add your LUKS keyfile into the initrd because that is the only accessible fileystem after GRUB boots the kernel and tells it to load the initrd into its memory space.
Asked in #grub. /boot is only accessible to grub for the initial boot, after that only the files provided within the initrd are available. Any suggestions for the best way to do this? From what I understand I would have to modify the Slackbuild to add a provided lukskey into the package, and the init file so it knows to use the file for cryptsetup luksOpen.
No you do not have to update the SlackBuild and the package.
What you need to do is run the command “mkinitrd -c” once. Afterwards there will be a directory tree /boot/inintrd-tree .
In that directory tree you make all the modifications you want.
You end with running the command “mkinitrd” (without any additional parameter). The mkinitrd command will compress directory /boot/initrd-tree into a new /boot/initrd.gz file which GRUB will pass to the kernel.
In the /boot/initrd-tree you need to:
– copy your LUKS keyfile to some directory below /boot/initrd-tree/ ; this directory is what you will use in the next step.
– edit the script ./init ; it is the initialization script where you will find the cryptsetup calls that use a keyfile to unlock the root filesystem. You need to make sure you change that call to use the path to your own keyfile (which you stored in /boot/initrd-tree so that it will be embedded in your new initrd.gz image).
This is just thinking out loud. I believe this is how it should work, but it is unproven.
Success 🙂 thank you.
Thanks for everything.
Happy Birthday!
> Mon, 16 May 2016 18:53:44 GMT
> veracrypt: updated to 1.17.
Thank you very much!
Love your work. I just upgraded to slackware64-current (soon to be 14.2) Do you have the multi-lib package for that? Let me know if you need help building them. I’d love to get my hands in there.
Richard did you really not look at all???
Hi. Was just looking around the applications on the Plasma 5 KDE. I tried Plasma Media Centre and it takes over the entire desktop. No way to exit so I had to use ctrl, alt, esc to kill it.
Is this normal or have I done something wrong somehow?
Thanks in advance for your answer
Nitesky there’s always Ctrl-Alt-BackSpace to kill a running X session, you do not have to do a complete reboot.
Other than that, I never used Plasma Media Center myself and only provide it as a package. When I started it, I could not see an exit button or configuration menu but simply pressing a keyboard key brought up the krunner dialog and made the task bar visible where I was able to close the window.
WOW! Ctrl-Alt-BackSpace killed the entire x-session as you said.
Ctrl-Alt-Esc killed only the Plasma Media Center and my x-session kept running.
Thanks for the tip
Hi, I installed version 1.01 of the xfce version of slackware64 live and it booted. But It’s asking for a user and password which I don’t have, and couldn’t find anything about it in the README. What am I missing.
From the README.txt and it is also mentioned in every article I wrote about liveslak on this blog:
Nice site alienbob.
Just reporting in your wordpress needs some patching. Getting popups from lqibh.com when browsing the site.
Hi Eric, I’m getting error with iso2usb.sh
./iso2usb.sh -i /images/latest/slackware64-live-current.iso -o /dev/sdb
(after about 10-11 minutes of file copying)
–Making the USB drive ‘/dev/sdb’ bootable extlinuz
/sbin/extlinux: cannot find device for path /mnt/alienusb.8zbLiY/boot/extlinux
*** ./iso2usb.sh FAILED at line 561
-another eric
Eric W, weird.
Is your USB stick large enough to contain the 2.6 GB of data in that Live ISO?
The error would indicate that the script has not renamed “/boot/syslinux” to “/boot/extlinux”, OR that the /boot directory was not or incompletely copied off the mounted ISO image.
What do you see if you mount the 3rd partition of that USB device and look inside the directory “/boot” ? There should be a subdirectory “/boot/extlinux” if all went well.
You can re-run the script with this commandline which will produce a LOT more feedback, I hope you are able to interpret it:
Hello Eric,
Are you planning on updating your VLC build to 2.2.4 or will you wait for 3.0 ?
I was originally only having visual artifacts playing *.mkv files with an intel HD 4000 in my laptop, but was able to fix that by increasing the buffer sizes. After building myself a computer with an nvidia graphics card I can’t seem to play any H.264 videos without them. Reading your post for the 2.2.2 release I installed mplayer/smplayer, and they were able to play anything wonderfully. I’m hoping a future version of vlc fixes these issues.
Thanks
Hi Eric,
is it possible to “sign in” somehow to your site so that I do not have to fill out name and e-mail to put comments every time?
You can not sign in to this site. You need to enter name and optional email/website every time.
hi eric,
for a printer to work i need the glibc-solibs from your multilib packages, and i just wonder if i only replace the slackware standard package of it with that one it would not cause any issue with my pure 64 bit?
Hi bilbo, you can replace the Slackware glibc-solibs package with its multilib version without penalty. But if you have other glibc packages installed as well, you should replace all of them with their multilib versions.
There’s no need to install multilib versions of gcc unless you want to *compile* a 32bit binary, and there’s no need for any of the “-compat32” packages unless you are using a proprietary 32bit program that requires a 32bit Slackware library.
Dear Eric,
I wanted to ask you a question re:Other work world IBM.
I believe you able to circumnavigate with Slackware at the Corp Office.
Can you elaborate if you are able how you coaxed this circumnavigation in favor of Slackware as your work laptop.
Applications and what not. (Calendaring, Corp Email, Corp GPolicy, Tools, etc.??
Maybe you have already mentioned this somewhere but I am really interested, for practical matters for me.
Source:
http://www.mylinuxrig.com/post/100003933891/the-linux-setup-eric-hameleers-slackware-linux
Warm regards
Hi _metic
Nowadays, IBM uses RHEL internally on employee computers with a layer of business applications on top of it. These internal IBM applications are witten so that they work both on Linux and on the Windows 7 platform which is also still used.
IBM does not care what you run on your laptop, as long as you adhere to all security guidelines, you install the corporate asset management tools and deliver the work they pay you to do.
In the past, I collaborated with other people at IBM who were Slackware fans like me and we made Slackware packages available to be able to do our daily job with Slackware. At that time, Windows XP was the only officially supported client platform even though a lot of people were already working on the creation of an internal Linux platform. I dual-booted Windows and Slackware.
Nowadays, I just use the RHEL7 installation for work, with Slackware running in a virtual machine.
Thank you for your response!
Hi Eric… First of all, thanks for your efforts.
Downloaded and installed latest kde5-packages on a fresh
64–current machine.
The question: is there a way to have the “Delete” button back in contestual menu in Dolphin?
I’ve searched but foud nothing, both in Kde docs and the net. I know that using “Shift+Del” makes the job but for me is better rightclicking the mouse.
Any suggestion?
Thanks again
HP
Hubert Phava, no idea what functionality you are talking about, never use graphical programs to delete stuff.
If some functionality in KDE has gone missing, the best thing to do is open a KDE bug report. And search the bugtracker first; perhaps someone ran into the same issue and already reported it.
Hubert Phava, next time post your questions in the relevant article, for instance the Plasma5 article I wrote yesterday and which probably made you download and install them.
I did your homework for you. This time only.
Go to Dolphin’s “Control” menu and select “Configure Dolphin…”. Click on “Services” in the left pane and scroll to “Delete” in the right pane. Add a check in its checkbox and click “OK”.
Now when you right-click in Dolphin there will be a “Delete” choice in the context menu.
TY Eriic and i apologize for my wrong posting… 🙁
> hu Jun 16 09:39:25 UTC 2016
> qt5: updated to 5.6.1 for Slackware-current (depends > on libxkbcommon which you will find in the same
> repository).
> qt5-webkit: updated to 5.6.1 for Slackware-current
> (depends on qt5)
As qt5 is not part of -current, I guess you are saying this is required if someone is using your KDE-5 packages in -current?
You are mixing up two separate repositories.
If you are using my KDE 5 (Plasma5) packages you will already have installed my qt5, qt5-webkit and libxkbcommon packages since they are part of its “deps”.
These packages that got an update in my regular repository are meant for those people who do not care for Plasma5 but do need Qt5.
Some users have run into issues as reported at ##slackware because of this:
$ md5sum slackware-current-install-dvd.iso MD5SUM
34d952b4312d98122e48a73d41337e0b slackware-current-install-dvd.iso
e334ab7590a70c137298bd44aa78f1e8 MD5SUM
Can you please fix it?
Source: http://bear.alienbase.nl/mirrors/slackware/slackware-current-iso/
Same here though I got a different md5sum on the .iso for some reason
3983a9925dc81c520df6cda5a19406e5
orbea explain to me what I should learn from your post?
This is how you should test the md5sum:
Well I tried to install packages from the iso in a chroot and it said the tarballs were corrupt. Then I tried installing it in virtualbox and on the installer it wouldn’t let me setup any partitions at all. I selected the option to start partitioning on the main menu of the installer and it just loaded it right back to that.
MachinaeWolf your MD5SUM is incorrect so your download is corrupt. That is the cause of all your issues.
And orbea could calculate the correct MD5SUM so his downloaded ISO is actually OK. However he used the wrong command and therefore made the wrong assumptions.
I downloaded it via wget and wget -c says it’s got the whole file. I’ll try again then…
Sorry, I did not understand how to check the MD5SUM, thanks for the correction.
$ md5sum –check MD5SUM
slackware-current-install-dvd.iso: OK
slackware-current-mini-install.iso: OK
Ok well I just did that too and it’s fine but my chroot I’m trying to make doesn’t want to install packages. I mount the iso and then cd to the mounted location and then cd slackware and do
$ installpkg –root ~/tmp/chroot/ a/*.t?z
and I get a load of errors like
Verifying package zoo-2.10_22-i486-1.txz.
/sbin/installpkg: line 378: /home/todd/tmp/chroot/var/log/setup/tmp/tmpsize9510: No such file or directory
/sbin/installpkg: line 378: /home/todd/tmp/chroot/var/log/setup/tmp/tmplist9510: No such file or directory
Unable to install a/zoo-2.10_22-i486-1.txz: tar archive is corrupt (tar returned error code 1)
orbea – your files check out, excellent!
MachinaeWolf – there’s the off chance that you have been downloading the ISO while it was just being re-generated. The server regenerates the ISO whenever an update to the ChangeLog.txt is detected (this check occurs once every hour).
Well it seems to like to take 2 hours to download at least every time…
Oh I realized my error… I mounted the iso to the chroot directory I was trying to install the packages to. It’s working now that I have that sorted out.
What’s your download speed MachinaeWolf ? Two hours is awfully long for a 2.7 GB file.
> Comment from alienbob
> Posted: June 17, 2016 at 23:04
> What’s your download speed MachinaeWolf ? Two
> hours is awfully long for a 2.7 GB file.
Just FYI, but since it was setup, your “bear” mirror has been consistently and considerably slower than other mirrors I’ve used.
Then it is good that there are more mirrors than just bear 😉
Ninety nine percent of the time I have no trouble or learning curve with slackware even areas that i would assume would be quite a challenge like installing virtual box is Ok because it only involves a build script , source and then installpkg command. Recently though concerning PHP which is 5.4.20 (cli) (built: Oct 11 2013 20:41:59) on 14.1 is not enough for some stuff i want to play with like Zend framework that needs 5.5 Now in /var/log/packages i can see php-5.4.20-i486-1 so i naively thought i should be able to upgrade & then deps phpadmin . It doesn’t look like being that simple . Assuming i’m not interested in having a LAMP stack but just need my core PHP upgraded to >= 5.5 whats the procedure?
ftp://ftp.slackware.com/pub/slackware/slackware-current/slackware/n/php-5.6.23-i586-1.txz
You should never mix packages from two different Slackware releases (-current is something entirely different from 14.1) because of the different libraries the packages link to.
Just build a new php package from the sources you can find in the slackware-current tree and you will end up with a package that works on Slackware 14.1.
cheers bob , when i ran $php -v it came up stating various libs missing & nothing clear on googling . I was about to just remove the install, then saw your post on slackpkg+ its seemed to have ungraded everything to current . bash-4.3$ php -v
//gives :
PHP 5.6.23 (cli) (built: Jun 24 2016 17:18:09)
and with a larvel in home , $php artisan serve works & gives no problem. So it looks Ok with desktop icons looking more modern etc
Eric,
I noticed that the Document Foundation has released Libreoffice 5.1.4 to patch a vulnerability in how it handles RTF files. The CVE entry for the vulnerability is here: https://www.libreoffice.org/about-us/security/advisories/cve-2016-4324/. A report from Cisco Talos with additional explanation is here: http://www.talosintelligence.com/reports/TALOS-2016-0126/
Hi John.
I have a big TODO…
Dear alienbob,
I am having various issues with PXE booting the most current version of Slack-ware live. After it boots the kernel, I get an error regarding Cg-manager and how it can’t find cgroup subsystems causing the computer to shut down. After messing around with some scripts in the iso image file to make Cg-manager non executable, the computer would still shut down regardless of the Cg-manager error being gone. Any wisdom or guidance would be much appreciated.
Best wishes,
Preston
Steamclient has issues with the latest mesa 12.0 on current.
libGL error: unable to load driver: swrast_dri.so
libGL error: failed to load driver: swrast
please ignore my last message. Error on my part.
Good afternoon.
Many thanks for the VLC update.
Just for general information, with the last couple of updates, and including this one, I get some flicker when playing DVDs. Doesn’t happen when playing a video file off the hard drive.
Also, when installing the last couple of versions of VLC, including this one, the following error is generated,
Executing install script for vlc-2.2.4-x86_64-1alien.txz.
Could not parse file “usr/share/applications/winecfg.desktop”: Key file contains unsupported encoding ‘”UTF-8″‘
Generating VLC plugins cache data…
Package vlc-2.2.4-x86_64-1alien.txz installed.
Again, just FYI.
Thanks for all your work which is “…above and beyond the call….”
🙂
I hacked around make_slackware_live.sh and have a diff that adds command line options for a different username (something other than ‘live’), a different user password, and a different root password.
It’s at http://dogpoker.space/make_slackware_live.sh.diff and all comments are welcome. The connection may be slow but it’s a small file (less than 7k).
Love your work, Eric.
Thanks,
Jim
I fixed a few errors and omissions that slipped by MIdnight Testing. The currently posted diff has been tested and works like it should.
Jim
32 bit PepperFlash 22.0.0.209
is available again from:
http://www.adobe.com/go/getflash
Hola e instalado slackware 14.2 plasma vivo en mi pc pero que contraseña usa luego de la instalacion pues no doy con cual es saludos
hola Alejandro, prueba usar el usuario live con el password:
live
tambien esta el usuario root
con el password:
root
Saludos
peebee yes I will see if my future pepperflash packages can use that as a source. At least we will have a 32bit Flash plugin for Chromium again then (and for Firefox users who install the freshplayer plugin as well),
Preston I have noticed the cgmanager related errors here too but I do not know what to do about them. I know too little of cgroups to debug the issue.
They do no harm here so I let them for what they are. Your shutdown appears unrelated to cgmanager anyway, from what I read of your description so there must be something else going on. Can you see what is happening in the console right before the shutdown? Apparently the live init completed successfully and your issue is somewhere in the Slackware configuration (may still be something custom I added to Slackware Live).
Hi jim jowski – that’s the good thing about open source… you can change all the bits 🙂
However I will not apply your diff to the liveslak sources. The username “live” is immutable as far as I am concerned – this is a Live Edition anyway.
People who re-generate the ISO from scratch and want different values for the ‘live’ and ‘root’ passwords should use the configuration file “make_slackware_live.conf” to specify those non-default strings.
alejandro, you were lucky that p431i7o answered your question…. this is an english blog for good reason: I do not speak or understand spanish (or portuguese for that matter).
Thanks p431i7o ! Perhaps http://docs.slackware.com/slackware:liveslak needs someone to write a proper translation!
Yeah, I’ll translate it to spanish.
Hello my question is I wanted to install live on my pc , slackware 64 kde 5, and I did the problem it is that the plasma wallet does not accept me I put passwords .But if I turn and pressed ctrl + alt + f2 me ACETA root passwords , the query is that why not allow me to start the greetings I am from Uruguay and I hope the translation and thanks for the previous reading as this is understood.
Hi alienbob,
I suggest you could include the package “breeze-kde4” in your ktown repo to introduce the breeze theme for qt4 apps
Thanks for all
Thanks for the reply, Eric. I understand your position fully. I can see where Slack-on-a-stick could be useful in a work setting which is why I did what I did. I can change the ‘diff’ to remove the command line options and leave in the variable changes to make_slackware_live.conf the place for the changes if you’d like.
lemin the “breeze” package in the ‘ktown’ repository contains the theme libraries for both KDE 4 and Plasma 5. No need for a separate package.
jim jowski, no need for a revised diff. You explained in proper detail why a useraccount other than the name “live” would be useful.
So I implemented a LIVEUID variable in the scripts (to be set in the .conf file and defaulting to “live”) and then looked at your posted .diff file. Almost identical except for the variable name – and the added commandline switches of course.
I added one more line: the skeleton files contain some ocurrences of the string “/home/live” and I fixed those too for the ${LIVEUID}.
These changes will end up in the git repository sometime soon, but I just started working on a new ‘ktown’ package set and that takes my time now.
alejandro – Google translate is not the best option to communicate… I do not understand if you have a problem with the Wallet password in Slackware Live, or after you have installed Slackware on your harddisk using the “setup2hd” script of Slackware Live.
Works for me! 🙂 I’ll patiently wait for your git.
Thanks,
Jim
I am enjoying experimenting with slackware live. It runs very fast on a USB3.0 — faster than the installed Win 10. That makes me very happy. You have done amazing work. Thank-You.
icewalker51 – nice.
hi eric,
I’m just curious about the missing underscore before alien tag in ur packages’ names 😀 … since it is somehow an unspoken rule, that everyone follows, and make ppl’s scripts for handling 3rd parties’ packages look prettier and robust! 😉
I know it is too much to ask, and you will probably kill me for asking this, but you are for sure master of those tiny scripts for changing something this little and make many of us lazy consumer happy! :))))
cheers
Eh, “it is somehow an unspoken rule, that everyone follows, and make ppl’s scripts for handling 3rd parties’ packages look prettier and robust” ?
There is no such rule. Everything is OK as it is.
Eric,
Is libva still a dependency for your wine package?
Jim
jim – I forgot to add libva to my virtual machine when compiling wine for Slackware 14.1 and that is why that dependency is not needed now. Obviously that is not what I intended, since the lack of libva means you will not have GPU video acceleration in Wine.
I will re-add libva as dependency for the Slackware 14.1 build next time.
On Slackware 14.2/current the libva package is already part of the distro so that was properly picked up by the wine package for those releases.
Hi,
Maybe I did something wrong… I’ve noticed a little issue with the Slackbuild for firefox localization : I had to provide VERSION=45.2.0esr to get it able to download the xpi file, but then the filetree inside the build package became “/usr/lib64/firefox-45.2.0esr” instead of “/usr/lib64/firefox-45.2.0”
In fact releases.mozilla.org/[…]/45.2.0/[…] isn’t a valid url but releases.mozilla.org/[…]/45.2.0esr/[…] yes so using VERSION=45.2.0 won’t make the automatic download working.
I had to 1) download the esr xpi 2) rename it to 45.2.0 (instead of 45.2.0esr) 3) build package using VERSION=45.2.0
Regards.
Moun, the SlackBuild was not written with Extended Support Releases in mind.
You found a way to create a working package for Firefox ESR using some minor adjustments, so there is nothing “wrong” in what you did.
Ok thanks for your answer. I’ve just understood that this SlackBuild was written for Slackware *current*. ESR type is only included in “stable” versions…
Will multilib ever be folded into the Slackware official packages? Have you written up the steps to build the multilib packages? I would try to do it but not sure what’s involved. Would it be possible to package up the environment to make multilib packages? Then maybe those packages could be included with official Slackware and those that need it can install the multilib-build package and build multilib for themselves.
Is this a boat load of work?
R.H. multilib will not be added to the Slackware distribution.
You can grab my packages or build your own. The process is written down in the README. All you need is an existing set of glibc and gcc multilib packages and install those. Download my glibc-multilib.SlackBuild and gcc-multilib.SlackBuild scripts and copy those into the official source directories, then launch both scripts and wait for new multilib packages to appear.
For the other “compat32” packages you can use my massconvert32.sh script that generates those from the official 32bit Slackware packages.
Have you actually read the documentation?
I did read the doc at the top of the site where the multilib lives. But I guess I failed on the comprehension. Thank you for taking the time to set me straight and sorry for that.
I’m going to have to make room to build all this on my system as I suspect it will need many GBs of storage.
I just updated to your release of the multilib packages and I am having problems with stddef.h. Not sure what’s going on yet.
Libreoffice-5.1.5 has a serious display bug which makes it virtually unusable:
https://bugs.documentfoundation.org/show_bug.cgi?id=101213
https://bbs.archlinux.org/viewtopic.php?id=215647
LO-5.1.4 seems to be OK.
So, it looks like https://cgit.freedesktop.org/libreoffice/core/commit/?id=d89abe0806947149eafbd9d7ce4b3095ec38b236&h=libreoffice-5-1 has to be applied to the 5.1.5 source code because otherwise we would be waiting for a 5.1.6 release until the end of october.
Hi
Feedback? do you guys REALLY care?
Let’s see then:
1- Make a official Bug-tracking system FGS! or create an official EMAIL for reporting bugs(not just for security bugs)
2- Clean the website from old/crap subpages
3-Replace Packages’s link from http://slackware.com/packages/ to http://packages.slackware.com. (whats the point of displaying OLD and useless page?)
4-Why not merge ALL packaing tools to one? like slackpm, spm, etc or something?
5- Merge ALL Slackware package/software repositories to one Central repository! (yours, slackbuilds.org, vbatts, http://rlworkman.net/pkgs/ and…)
Ofc everyone can have their own package center, but not SW team (IMO)
GL
Alex, please note that I am not affiliated with Slackware, Inc. Your rants will have to be addressed at info@slackware.com.
In the meantime, I can give you some answers, because it is good that you care.
(1) A bug tracker requires (human) resources which the small team of one person (plus a few volunteers like me) can not spare. You have issues, spill them on the linuxquestions.org forum like everyone else does, and you will be heard.
(2) It may happen during this development cycle. Again, it is a matter of prioritization. For up to date information on Slackware, please visit http://docs.slackware.com/
(3) Would be addressed by (2)
(4) There are several 3rd party tools that are not part of Slackware (the ones you mention I do not recognize so they are surely 3rd party) and other than that, they are integrated. slackpkg works by calling the various components of the pkgtools. And there’s no more to it than that.
(5) There is a central repository only for Slackware itself. The rest is not “official”. Me, Robby, Vincent, we are all not getting any money out of this, we do it as a hobby, for fun. Either of us can just stop releasing packages or even abandon Slackware if “real life” became more timeconsuming or more important, and who would then maintain our packages in the “central repository”? Again it is a matter of resource prioritization. We leave Pat to keep the core distro running while the rest of us creates the dressing.
(6) As for your challenging and condescending tone, it does not become you. You have issues with Slackware? Deal with it or move on.
Excellent response, alienbob!!
I installed the last KDE 5_16.08_testing
But i get this error in configuring skype account:
<<>>
And google hangout not working…
Thank you
If you install a “testing” release of mine, I really do expect that you know what you are doing. And actively try to find solutions for your issues instead of letting me guess what’s wrong on your computer. Please install the ‘latest’ release instead.
all works fine
unless instant messaging
Hi Eric, I installed KDE 5_16.08 latest. So far everything works great. Thanks!
I’m running Slackware 14.2 with the provided NetworkManager 1.2.2-x86_64 on an AMD A8-7410 APU (Dell Inspiron 15 – 5000 Series).
For one reason or another I lost the ability to browse the internet and, suspecting malware, formatted the partition and reloaded the O/S. The problem persisted, turning out to be a DNS issue (possibly related to the ISP intercepting/redirecting DNS requests).
If I put the nearest router IP in /etc/resolv.conf, that worked until the next time I rebooted, then it came up with information varying from being updated by Network Manager to eth0.dhcpd. In no case did it have the DNS servers ip numbers. If I manually populated it with the IP of the nearest upstream router, it worked until I rebooted, then, not so much.
Believing that dnsmasq could supersede NetworkManager and dchpd with hard-coded DNS server addresses, I enabled /etc/rc.d/rc.dnsmasq and opted in the /etc/dnsmasq.conf file as follows:
no-resolv
no-poll
server=209.244.0.3
server=209.244.0.4
That did not get me DNS lookups. Instead, something was trying to do DNS lookups on the loopback interface so I opted for:
interface=eth0
… in the /etc/dnsmasq.conf file but that also had no effect. So, noticing that the contenst of /etc/resolv.conf sometimes came from /var/run/NetworkManager/resolv.conf, I tried changing that but it also changed as I rebooted but I figured there was a way to configure Network Manager. Looking in /etc/NetworkManager I found a pointer in NetworkManager.conf to /etc/NetworkManager/conf.d but that was about all I found there.
/etc/NetworkManager/conf.d/00-dhcp-client.conf offered the ability to change the DHCP client from the Slackware default but what would be the trickle-down consequences of that?
I’m a newbie and trouble-shooting is an involved process.
Ultimately, I solved the problem by adding:
dns=dnsmasq
…to the /etc/NetworkManager/NetworkManager.conf file.
I’m guessing you have input on the ultimate densign and release info of Slackware. I don’t see that documented anywhere in Slackware but I’d like to suggest a helpful documenting in the NetworkManager.conf file, i.e.:
# Uncomment the following line if you have /etc/rc.d/rc.dnsmasq enabled and dnsmasq.conf configured.
# dns=dnsmasq
That may be obvious to you and there may be some documentation on NetworkManager or Slackware I’m overlooking but maybe some other newbie would benefit from my experience just by having that comment in the .conf file.
As for the original problem that caused me to reload the O/S, it seems the DNS replies were coming back to the router from TestNet3 (203.0.113.0/24). I don’t know why. I block those kinds of IANA IP ranges, at all junctures, owing to their reputation for being associated with attacks.
Anyway, this fix works. Hope it helps someone else.
Hi John,
The issue you describe is not specific to any customization I offer for Slackware, but since this is the feedback section I have an answer for you.
You enabled dnsmasq to overcome an issue with your local ISP but dnsmasq does not “do” anything by itself – you have to actively use it just like any other service. So you found the correct way to let NetworkManager query dnsmasq as a DNS resolver by adding “dns=dnsmasq” – good!
Not every user-contributed hint and information needs to go into Slackware’s core though – that puts an unnecessary burden on the maintainer (Pat). That is why we now have the Slackware Documentation Project where you can become a contributor to the shared body of knowledge about the Slackware distribution and its use.
I know the fix is simple but this is driving me crazy. I’m building a custom iso for slackware64-current using make_slackware_live.sh and it builds. When I try to put it on a USB stick I get the following:
— Making the USB drive ‘/dev/sdi’ bootable using extlinux…
/mnt/alienusb.k1nthW/boot/extlinux is device /dev/sdi3
rsync: mkstemp “/mnt/alienus2.BiepOV/EFI/BOOT/theme/.index.html?C=D;O=A.TFxi0r” failed: Invalid argument (22)
rsync: mkstemp “/mnt/alienus2.BiepOV/EFI/BOOT/theme/.index.html?C=D;O=D.7NtYks” failed: Invalid argument (22)
rsync: mkstemp “/mnt/alienus2.BiepOV/EFI/BOOT/theme/.index.html?C=M;O=A.zFNEFs” failed: Invalid argument (22)
rsync: mkstemp “/mnt/alienus2.BiepOV/EFI/BOOT/theme/.index.html?C=M;O=D.pOll0s” failed: Invalid argument (22)
rsync: mkstemp “/mnt/alienus2.BiepOV/EFI/BOOT/theme/.index.html?C=N;O=A.RCn2kt” failed: Invalid argument (22)
rsync: mkstemp “/mnt/alienus2.BiepOV/EFI/BOOT/theme/.index.html?C=N;O=D.GuDJFt” failed: Invalid argument (22)
rsync: mkstemp “/mnt/alienus2.BiepOV/EFI/BOOT/theme/.index.html?C=S;O=A.05er0t” failed: Invalid argument (22)
rsync: mkstemp “/mnt/alienus2.BiepOV/EFI/BOOT/theme/.index.html?C=S;O=D.F9k9ku” failed: Invalid argument (22)
rsync error: some files/attrs were not transferred (see previous errors) (code 23) at main.c(1178) [sender=3.1.2]
*** ./iso2usb.sh FAILED at line 691 ***
— Cleaning up the staging area…
I can’t find the offending files using locate.
Jim
I was able to find the offending files (they are created by ‘wget’) and remove them. What now puzzles me is the differences between the EFI directory in http://bear.alienbase.nl/mirrors/people/alien/liveslak/ and http://mirrors.slackware.com/slackware/slackware64-current/. Bear has EFI/BOOT/theme and the other one doesn’t. Is the theme directory really needed?
Jim
Hello, Eric.
In Slackware64-14.2 cdrecord does not write DVD disks from regular user but only from root despite I have included my user into cdrom group. It shows next errors:
$ cdrecord dev=/dev/sr0 slackware64-14.2-install-dvd.iso
Cdrecord-ProDVD-ProBD-Clone 3.01 (x86_64-unknown-linux-gnu) Copyright (C) 1995-2015 Joerg Schilling
cdrecord: Operation not permitted. Warning: Cannot raise RLIMIT_MEMLOCK limits.
cdrecord: Cannot allocate memory. WARNING: Cannot do mlockall(2).
cdrecord: WARNING: This causes a high risk for buffer underruns.
cdrecord: Operation not permitted. WARNING: Cannot set RR-scheduler.
cdrecord: Permission denied. WARNING: Cannot set priority using setpriority().
cdrecord: WARNING: This causes a high risk for buffer underruns.
cdrecord: Insufficient ‘file read’ privileges. You will not be able to open all needed devices.
cdrecord: Insufficient ‘file write’ privileges. You will not be able to open all needed devices.
cdrecord: Insufficient ‘device’ privileges. You may not be able to send all needed SCSI commands, this my cause various unexplainable problems.
cdrecord: Insufficient ‘memlock’ privileges. You may get buffer underruns.
cdrecord: Insufficient ‘priocntl’ privileges. You may get buffer underruns.
cdrecord: Insufficient ‘network’ privileges. You will not be able to do remote SCSI.
scsidev: ‘/dev/sr0’
devname: ‘/dev/sr0’
scsibus: -2 target: -2 lun: -2
Warning: Open by ‘devname’ is unintentional and not supported.
Linux sg driver version: 3.5.27
Using libscg version ‘schily-0.9’.
Device type : Removable CD-ROM
Version : 0
Response Format: 1
Vendor_info : ‘ASUS ‘
Identifikation : ‘SDRW-0806T-D ‘
Revision : ‘AE07’
Device seems to be: Generic mmc2 DVD-R/DVD-RW/DVD-RAM.
Using generic SCSI-3/mmc-3 DVD+RW driver (mmc_dvdplusrw).
Driver flags : NO-CD DVD MMC-3 SWABAUDIO BURNFREE
Supported modes: PACKET SAO LAYER_JUMP
cdrecord: Warning: Cannot read drive buffer.
cdrecord: Warning: The DMA speed test has been skipped.
cdrecord: Operation not permitted. rezero unit: scsi sendcmd: fatal error
CDB: 01 00 00 00 00 00
cdrecord: Operation not permitted. Cannot send SCSI cmd via ioctl.
cdrecord: Operation not permitted. Cannot open or use SCSI driver.
cdrecord: For possible targets try ‘cdrecord -scanbus’. Make sure you are root.
cdrecord: For possible transport specifiers try ‘cdrecord dev=help’.
Best Regards,
Anton.
Hello, Eric.
I have solution of this problem. I just did next steps:
# ls -l /usr/bin/cdrecord
-rwxr-xr-x 1 root root 480232 May 31 22:44 /usr/bin/cdrecord
# chown root:cdrom /usr/bin/cdrecord
# chmod u+s /usr/bin/cdrecord
.
Now cdrecord works fine from regular user. And i think it would be nice if you have included these strings into the doinst.sh script of cdrtols package, of course, if you find this solution to be acceptable.
Best Regards,
Anton.
Anton, I recommend reading the CHANGES_AND_HINTS file located in the root of every Slackware release. For Slackware 14.2 this file is located here: http://ftp.osuosl.org/pub/slackware/slackware64-14.2/CHANGES_AND_HINTS.TXT
Your issue with cdrecord is covered in that file.
jim jowski, I think you are confusing two different things.
Liveslak creates a Live Edition of Slackware. For this, you need the directory “liveslak” plus everything below. You can download that directory tree from “bear” or any mirror but using “wget” for recursive downloads is probably the worst of all available options. Use “lftp” or “rsync” instead please.
The directory you mentioned: “EFI/BOOT/theme” is required by liveslak. It has nothing to do with the original Slackware. Therefore it is not present on an official Slackware mirror repository.
Hi Eric
Many thanks for Chromium-53 for 14.2 and current.
Will updates to the 14.1 version now freeze at Chromium-52?
Thanks
peebee
peebee – I only create packages when I have time for that. And currently I need my time for Real Life issues.
So I can not answer your question.
Eric,
The more I use liveslack; the more I learn. 🙂 Those index files were aren’t removed when the “-f” option is used with make_slackware_live.sh. Now that I’m aware that both ‘lftp’ and ‘rsync’ can be used; I’ll use them. As for ‘wget’: old habits die hard. Twenty years of doing something is hard to change but change in this case is good.
Thanks,
Jim
Hi Eric,
The latest chromium package has a mixup with keyrings. In Plasma, upon starting chromium, there is a dialog that tells me that a “default keyring” needs to be accessed. And then, *some times* the KWallet dialog appears. At all times, chromium does not get my passwords and I have to relogin every time to my sites.
Thanks,
Eduardo
Hi Eduardo,
I noticed that kwallet popped up with a request to create a new wallet when I started the upgraded chromium. But, I can not find any new wallet and all the passwords stored by chromium are still available in the browser and used to logon to sites automatically.
Eduardo wrote:
> The latest chromium package has a mixup with keyrings. > In Plasma, upon starting chromium, there is a dialog that > tells me that a “default keyring” needs to be accessed.
Yes, I’ve had the same problem when using chromium to log into Netflix. I figured out how to use widevine with Vivalidi and have not had the same problem, so it looks to be something, as Eduardo said, to do with chromium.
I’ve been a long time user of your slackbuilds. Sorry to hear that you are getting laid off. I recently retired from California to Japan, and was fortunate enough to do practically the same job for 33 years, but going from contractor to contractor. Best wishes finding a job soon. Your background and knowledge is far greater than mine.
Sid
Sorry to hear about you getting laid off and sorry for a non-technical question, I have made up my mind to start a career in IT. What IT technology/subject/certificate would you recommend for someone starting a career in IT for a relative long-term period; since most of the IT subjects available or recommended by the current academia are out of date/or would be out of date in the very near future. Your perspective from someone who has learned, worked and again needs to relearn would be very valuable for me.
Thank you
Bjorn I can not comment on that. It would depend in what area of IT you want to pursue a career. If you are going to work as a network engineer, then a Red Hat certification does not make sense. If you want to make programming a part of your career, then courses in Java, Python and such would be an option. But ultimately, there is no good answer I can provide you with.
In reference to the problem reported by Eduardo a few messages back were
he said,
> The latest chromium package has a mixup with keyrings.
> In Plasma, upon starting chromium, there is a dialog that
> tells me that a “default keyring” needs to be accessed.
While I was using Vivaldi-1.3.551.38, this was not a problem. However I have just
upgraded to Vivaldi-1.4.589.11, and the problem has appeared just as Eduardo described it. According to the “help” content from the pull down menu, this new version of Vivaldi is based on chrome/53.2785.90/blink53736. So, as they say,
this problem appears to from “upstream.”
Just FYI.
AMD could use a good man with your great abilities
The problem mentioned two post back regarding “default keyring” has disappeared with the release of
Vivaldi-1.4.589.29.
Then I assume that the issue will also disappear with the next release of Chromium. I still have to type two passwords to load my chromium passwords and preferences
Hello, Eric.
I am a 62 year old Italian Linux user, first with Red Hat, up to version 7.2, then inevitably with Slackware since 2001.
First of all I want to thank you for the great contribution you are giving to the development and maintenance of Slackware, and my compliments to the site your personal and Slackbuilds.org.
I have a question to ask.
I would like to create me my personal repository compiled Slackware packages (the ones that I keep in my hard drive dedicated to it) and to create packages.txt files using the script of slapt-get; but despite having read dozens of web pages, I have looked everywhere, I do not know how to generate the MANIFEST.bz2 file.
I tried to use your gen_repos_files.sh script but, unfortunately, is too complicated for my knowledge of the bash script and, above all, ask me generation and the keys used for encryption that I never use.
I thank you right now if you have time and desire to help me or suggest me where to find the instructions for creating a personal website compiled packages.
Sorry if I express myself badly, but I know little English and I relied on Google translator.
Ferrari Renato Maria
Hi Renato,
The script “gen_repos_files.sh ” has a parameter “-h” which will display some helpful text. It should not be too hard to use even for people without shell scripting skills.
If you do not have a GPG key, the script will warn that it will not be able to generate GPG signature files but it’s only a warning which you can safely ignore.
The one thing you are required to do before using that script is to run the command “uuidgen -t” and copy the resulting UUID string into the script’s “RSS_UUID” variable. If you are not sure you can do that, there is an alternative, which is to create a configuration file which the script will always use when it runs,
Create a new file in your homedirectory and call it “.genreprc”.
In that file, copy the following lines, and make sure that you write something sensible inbetween the empty “” strings:
REPOSROOT=””
REPOSOWNER=””
RSS_UUID=””
The REPOSROOT variable must contain the directory where you keep your package repository.
The REPOSOWNER variable should contain your name and optionally your email address. It will be used to search for a GPG key and it will be used in the repository metadata files.
And finally, RSS_UUID variable must contain the output of the command “uuidgen -t”. This is a string which uniquely defines your repository.
You can add more variables to the “.genreprc” file but they are all optional.
When you have created this file, you can run the script and that should generate the MANIFEST and other files.
Hi Eric! You plan to build Plasma 5.8?
> Sun Oct 23 10:24:48 UTC 2016
> veracrypt: updated to 1.19.
Thank you very much. Greatly appreciated!
Hi Eric,
I’ve been using your vlc binaries for some time and had great experience with them so far. Today I’ve setup a 64bit Slack v14.2 and installed your latest restricted build of vlc (vlc-2.2.4-x86_64-1alien.txz). It worked great until I needed to use phantomjs which is dependant on libfontconfig. I’ve installed the official slackware64/x/fontconfig-2.11.1-x86_64-2.txz and since then vlc breaks with Segmentation fault. Removing fontconfig fixes the issue.
It will be great if can have a look. Thank you!
-Ivan
Dear Eric
on the MANIFEST file: thanks for your valuable suggestions: the program works fine.
The only small flaw was due to the fact that I had copied and pasted the instructions directly from this page; but ” of the web page are different from those of the bash script, correct my mistake, I resolved.
Thanks to script of Patrick Volkerding , of Eric Hameleer, to the Advanced Bash Guide, (and to my friend Emilio Conti that has translated into Italian), the Italian slackware http://www.slacky.eu community and all people who have posted on the web their script, I know a little ‘in bash programming; so I will try, as well, to automate the generation of my personal repositories.
thanks for the updated veracrypt package.
Hi Eric, just wanted you to suggest the addition of two packages to your KDE5 collection: KSnip and Krusader.
You may already know Krusader and KSnip is a tool very similar to the Win10 Snipping Tool
GitHub homepage:
https://github.com/Ssiws/KSnip
For Krusader, simply editing the SBo SlackBuild is what it takes to build it.
For KSNip, adapting any KDE app SlackBuild also works. In any case, I have SlackBuilds who build correctly both apps and I would be delighted to share them with you should you like to take a look at them.
Thanks again!!
I forgot to say:
I suggest Krusader because there is a new version (2.5.0) which is KDE5 only.
I suggest KSnip because I find the tool extremely useful, even better than Spectacle for a quick screen snip capture.
Thanks!
Sorry agaiin. Ksnip’s GitHub page was incorrect. Here is the correct one:
https://github.com/damirporobic/ksnip
There is also the linux-apps page:
https://www.linux-apps.com/p/1156408/
Apologies for the mistake.
Eduardo,
There’s lots of packages that work well with Plasma 5, but I am not going to add any or all of them.
Dolphin and Spectacle are the official file-manager and snapshot tool, and I leave it at that.
> Sun Oct 23 10:24:48 UTC 2016
> veracrypt: updated to 1.19.
With the previous version it was not possible to close a “container” and it was necessary to terminate the program, which, in turn, closed the “container.”
With the latest version terminating the program, does, of course, close the program, but the container remains open. Now the only way to close the container is reboot the computer.
Just FYI.
cwizardone if you find a fix let me know.
However, I have personally not experienced what you are describing. For me, dismounting the volume cleans up; no device-mapper entry remains.
Hi,
hexchat is available in Slackware 14.2 / current. If there is not a special reason to keep it, maybe it is a candidate to be removed from the corresponding repositories?
Note: It appears in my slackpkg++ packages-to-be-upgraded list, that is why I inform you.
martinowar yes I should take care of that.
Calibre 2.71 has an error
This application failed to start because it could not find or load the Qt platform plugin “xcb”
in “”.
Available platform plugins are: eglfs, linuxfb, minimal, minimalegl, offscreen, xcb.
Reinstalling the application may fix this problem.
Aborted
I’ve had this error in the past and have the libxkbcommon package.
Post on linuxquestions about this from earlier this year.
https://www.linuxquestions.org/questions/slackware-14/calbre-fails-to-start-4175574668/
Hi boris.
Is this on Slackware 14.2 or -current? If you installed Qt 5.7.0 then apart from ‘libxkbcommon’ you also need to install the ‘libinput’ package.
-current, and was missing libinput.
Thanks
> Sun Oct 23 10:24:48 UTC 2016
> veracrypt: updated to 1.19.
> With the previous version it was not possible to close a “container” and it was
> necessary to terminate the program, which, in turn, closed the “container.”
> With the latest version terminating the program, does, of course, close the program, > but the container remains open. Now the only way to close the container is reboot the > computer.
> Just FYI.
> Comment from alienbob
> Posted: November 8, 2016 at 20:48
> cwizardone if you find a fix let me know.
> However, I have personally not experienced what you are describing. For me,
> dismounting the volume cleans up; no device-mapper entry remains.
Apparently it has something to do with KDE-4(in current) or a KDE-4 application.
That thought ocurred to me, so I exited KDE, ran xwmconfig, picked Xfce, started
the same, and was able to successful close the container that would not close while
running KDE. Don’t know if that makes any sense?
🙂
That matches with my experience… I only use veracrypt on XFCE. I will have a look at what happens on KDE when I have some time.
Another thought is it just might have something to do with Gwenview.
BTW, “Welcome back,” so to speak. 🙂
sorry to bother you Eric,
but is there any chance of getting chromium-54 on slackware 14.1?
I saw it on 14.2 on your builds and that the files: chromium-53.0.2785.92-x86_64-2alien.lst and chromium-53.0.2785.92-x86_64-2alien.meta where updated on november 18th. So I thought there may be an update close.
I ask this because of this bug that came with chromium-53.0.2785 NET::ERR_CERTIFICATE_TRANSPARENCY_REQUIRED https://bugs.chromium.org/p/chromium/issues/detail?id=665484&q=NET%3A%3AERR_CERTIFICATE_TRANSPARENCY_REQUIRED%20&colspec=ID%20Pri%20M%20Stars%20ReleaseBlock%20Component%20Status%20Owner%20Summary%20OS%20Modified that was supposedly solved on chromium-54
I never had a chromium 53 package for Slackware 14.1. It’s currently still at chromium-52.0.2743.116 because I could not get chromium 53 to compile on Slackware 14.1.
I can try again with the new chromium 54.
Well, this is weird https://postimg.org/image/lkw4091sl/ so this is a chromium 52 package?
I should have been more careful in my wording. The 32bit package could not be compiled successully so http://www.slackware.com/~alien/slackbuilds/chromium/pkg/14.1/ still has a chromium 52 package. But there is a 64bit package.
Compiling chromium 54 on 64bit Slackware 14.1 just failed, nevertheless. It looks like it still needs a patch I removed because it was not required on 14.2/current.
Regarding our discussion of Veracrypt a few messages back. After using Veracrypt this morning in KDE-4-in-current, it wouldn’t dismount a container, which is its usual behavior.
Just for chuckles I logged out of KDE, then started it again, with, of course, startx, and was able to close the container.
Just FYI.
p431i7o I think I finally created a correct patch that allows chromium-54 to compile on Slackware 14.1. If the compilation completes without further issues, I hope to have the packages (64bit and 32bit) available tomorrow.
Thanks Eric!
that are great news,
I hope in this December make finally my transition at home of my desktop to Slackware 14.2
but this will relieve a situation at home (wife having much problems with Firefox and its flash player plugin) to watch Videos.
Hi Eric, it’s me again with another nuisance.
I learned that for KDE Apps 16.12 KUser is going to be removed. With that removal there goes (AFAIK) the only graphical utility for user management in Slackware.
My question is: is there a possibility of keeping it from 16.08?
Thanks!
Hi Eduardo,
Looking at https://community.kde.org/Applications/16.12_Release_Notes you can see that 16.12 has a big change set. And indeed, kuser will no longer be shipped in source form together with KDE Applications. I guess it is not a big deal to keep a copy of the 16.08.3 version and include a package for that release. I do plan on releasing packages for Applications 16.08.3, in fact I hope to have a Slackware Live Plasma5 Edition (64bit) ready tomorrow with all the latest KDE packages, the official release in my ‘ktown’ repository will be somewhere during the weekend.
Thank you Eric!
Many thanks for your work Eric, Slaxkware is great, but only thanks to people like you.
oops, Slackware of course 🙂 …but it could be the name for Slax and Slackware “cocktail”
Merci beaucoup eric, god bless you
Hi Eric, this is just to tell you that -current latest update ships libinput-1.5 while the last KDE update yhou shipped has libinput-1.3. In case one has blacklisted the ‘alien’ tag in slackpkg, the box would stay with your older version of libinput and then X11 would start but be unable to receive any input. Users of Ktown should be aware that they should make sure their libinput is updated. Thanks!
Hi Eric,
Thanks for updating your logwatch package. I meant to give feedback on it a while back but always forgot to do so. Your current package overwrites all configuration files in /etc/logwatch/conf/. I can send you a patch if you want to fix this.
Thanks again for all your work.
Sure Artourter, patches are always welcome.
Since no one has requested these on this page, can I have slackbuilds for these two?
https://github.com/Airblader/i3
https://github.com/Airblader/i3blocks-gaps
Thanks in advance, Mr. Alien.
Hi azure,
I am not a fan of tiling window managers and would not be able to quality-check the packages if I would create them. Furthermore, I would not use them either. And I noticed that this is an unofficial fork or “i3” which is already available on SlackBuilds.org. Perhaps you could ask the maintainer of that i3 SlackBuild on SBo to create a script for your requests?
Hi Eric,
Is this tagfile still usable in 14.2?
http://www.slackware.com/~alien/tools/minimalslack/
anon, you tell me. I am pretty sure that “live_xfcebase.template” should be usable. Of the rest I am not certain because I have not checked them for years. Try your luck, tell me the results.
Hello,
i want to ask why I receive this message?
upgrade ktown breeze-5.8.4-x86_64-1alien –> breeze-5.7.3-x86_64-1alien
is this mean that ‘ktown’ is not updated?
Svetlyo you should post this kind of questions in the relevant thread – in this case, your question is about Plasma 5 packages for which there is a new article almost every month. Those articles already contain the answer to your question.
I warn against NOT using the “testing” repository since that is outdated. You must use the “latest” repository instead.
Please check the configuration of your package automation tool (slackpkg+ or slapt-get) and ensure you are using the correct repository URL. The package “breeze-5.7.3-x86_64-1alien” is only available in the “testing” repository.
Hi, just going through the motions again of upgrading to multilib for slackware64-14.2 – do you think Patrick will ever see the light and make it the default? There is no advantage in not providing this feature, and 32bit is not going away any time soon for one reason or the other (steam..)
He will not add multilib to slackware64.
Just connected to http://alien.slackbook.org/blog/
Seeing this: hacked by NG689Skw
Fix?:
https://www.wordfence.com/docs/how-to-clean-a-hacked-wordpress-site-using-wordfence/
Yeah, I restored the page, updated everything that needed updating and checked for changed files (none out of the ordinary). Thanks for spotting & reporting.
Hi,
I really appreciate all your work.
My question it’s founded in curiosity, so sorry if it’s not pertinent of your job (and sorry for my poor english).
Do you know ConnochaetOS?
https://connochaetos.org/wiki/home
I would like to know if it’s really hard to make a livecd (but installable too) following the same objective as this Slackware version has?
I installed once, I really liked it all, slim, light, simples.
But I would like they would have a 64 bit version.
Well, just by curiosity I’m asking.
Thanks for your effort.
Samuel
Hi Samuel,
I think what you want is possible with liveslak.
I looked at their repository https://connochaetos.org/slack-n-free/slack-n-free64-14.2/ and it seems they only offer a few (replacements for) packages that would come on top of a regular Slackware. However there does not seem to be any documentation as to what Slackware packages are replaced and whether there are any Salix packages that should be installed as well.
My liveslak scripts are based on the availability of packages in a directory structure. If the distro’s package set is too different from a regular Slackware installation then you can write one or more package list files (one for every repository that contains the packages you want to install into the Live OS) and the liveslak scripts will download a copy of the repositories and install the packages from that copy.
But you first should have an overview of what repositories are used, and which packages from what repositories are needed.
Hi Eric,
I ‘ve been using your calibre slackbuild since day one and had no problem till recent.
get this at prompt:
bash-4.3$ calibre
This application failed to start because it could not find or load the Qt platform plugin "xcb" in "".
Available platform plugins are: eglfs, linuxfb, minimal, minimalegl, offscreen, wayland, wayland-xcomposite-egl, wayland-xcomposite-glx, xcb.
Reinstalling the application may fix this problem.
Abortado
Reinstalling calibre don't fix the issue
I run a fresh Slackware 14.2 with multilib
using calibre 2.79.1
and deps installed using slackpkg:
libxkbcommon-0.7.1-x86_64-1alien.txz
libinput-1.5.4-x86_64-1alien.txz
podofo-0.9.2-x86_64-1alien.txz
qt5-5.7.1-x86_64-1alien.txz
qt5-webkit-5.7.1-x86_64-1alien.txz
thanks for the hard work on Slackware :) and any help is welcome
Atte.
César
Hi cesarion76 – in addition you now have to install the “libwacom” package, it is a new dependency of “libinput”.
Thanks Eric, Calibre is working again.
Shouldn’t be “libwacom” included on the description file: “calibre-2.79.1-x86_64-1alien.dep”?
Just wondering.
Thanks again!
Atte. César
Hi Eric, thanks for you response!
So, I could do something crude like list the packages that distro is using, replace on liveslak without modifing your directory structure and run to see what happen? Besides checking the repository area used, right?
I was checking that link you passed and find odd how few packages has.
So I remember they only make a 32 bits distro.
So, I should base the list of packages above that link: https://connochaetos.org/slack-n-free/slack-n-free-14.2/
and see if have the same packages on 64 bits.
Only problem could be the free-kernel he uses…
Well… anyway, just curiosity.
Thanks
Hi Eric. I already compile the “mscore-2.0.3”. Do you want that to share in your repository?
Thanks for qt5 !!! 🙂
Lindberg Luiz, I do not add other people’s packages to my own repository. If you want, you can contribute your SlackBuild script to the slackbuilds.org script repository.
I want to use a development snapshot of Scribus called 1.5.3.There are versions for many other Linuxs but not slack.
There is also an appimage file but i don’t know how to run it,
I do not see a question in there that’s relevant to me, John.
Hi Eric,
Thanks for all the work you put in. You are for sure one of my heroes.
Anyway I was looking at the new KDE5 packages you put online. I noticed that dolphin was compiled with libbaloowidgets included but wasn’t libbaloo deprecated ?
( I only checked the x86_64 )
I build dolphin again ( you build script is a amazing )
and now it is working fine
I just wanted to let you know. All the best to you and everything you hold dear.
Hi David
Indeed, I deprecated the kdelibs4 based “baloo-widgets” package. But in the set source tarballs for the KDE Frameworks 5 based Applications there’s also a “baloo-widgets”. This KF5 based one compiles into a package “baloo5-widgets” and that one is not deprecated. The library in that package is called “libKF5BalooWidgets.so”, not “libbaloowidgets.so”.
But I fail to see how you came to the conclusion that dolphin was compiled against baloo-widgets? It is not, in my opinion.
Dear Sir,
My PC runs Slackware 14.2 x64.
Today, after many years, I tried the Opensound system again, oss-linux-v4.2-2017-i386.tar.bz2. And I must say, I am impressed with the perceivable difference in sound quality, over the included ALSA driver. The music files, all lossless, have more perceivable detail; an instant upgrade of my system\’s soud output.
Alas, it seems that VLC does not have OSS support compiled in by default. Over a number of hours, I have tried and failed, to successfully compile VLC with OSS support.
Would it be possible for you to compile, and host on your site, a version of VLC which (also) includes OSS support?
Thank you for your time.
Sincerely,
Alex
Alexander I would have to dig into that, and I do not have the time or energy for that.
So, I hope you find out how to add OSS support to VLC yourself, without requiring another package to be installed onto the system (the ‘oss’ package in particular, because I do not want to add a new dependency that almost no one uses) and share your findings here.
I use Twitter Bootstrap for front end and beginning to dabble with LESS. So the process will be edit bootstrap.less & variables.less with geany by hand. The problem is finding something compatible with 32 bit Slackware14.1 Nodejs is on slackbuilds but tools like grunt don’t seem to be available. Any suggestion on how to convert LESS to CSS i’m reasonably OK on command line
andy, I have no clue at all about LESS, Node.js or Grunt. I am not a developer.
But reading https://gruntjs.com/getting-started I would think there is a reason that grunt is not found on SBo: “Grunt and Grunt plugins are installed and managed via npm, the Node.js package manager”.
Hi Eric,
thanks for all the great work you do with/for Slackware, especially the Plasma Packages.
I am using your ktown packages on top of Slackware64 14.2 current. I am having some problems with Kate and/or KDevelop. Whenever I want to start them, I get the error message:
kate: error while loading shared libraries: libKF5SyntaxHighlighting.so.5: cannot open shared object file: No such file or directory
Unfortunately I can’t tell you when this problem started, since I don’t use this machine everyday. Can you help me with this problem?
Christian – Slackware 14.2 *or* -current? It’s one or the other, not both.
You should base the selection of the ktown package repository on the answer to the above question.
Do you have the syntax-highlighting package installed (part of Frameworks) which contains that library which is mentioned in the error? Does a “ldd /usr/lib64/libKF5SyntaxHighlighting.so.5.32.0” show any libraries with a “not found” status?
Thanks for the quick reply, I am using “current”, my mistake. You were right, I was just missing the syntax-highlighting package.
Hi alien bob, i guess question about about LESS wasn’t really relevant in context of Slackware packages. However bootstrap use of LESS may become relevant to web sites as Google pushes for webs to be smart phone friendly etc. Anyway forget LESS twitter is moving to SASS. For anybody interested put use of SASS here: https://twitter.com/andybrookestar/status/853897374254739456
Hi alienbob,
Is it intentional that glibc-debug-2.25_multilib-x86_64-2alien is missing from your current multilib repo? I do not see anything in the changelog.
Orbea, the latest Slackware glibc.SlackBuild has disabled the creation of a glibc-debug package so I followed that example.
Do you actually use that debug package?
alienbob, I might need it now as gcc-7.1.0 is being rather problematic. For example it gets stuck indefinitely on a cc1plus command in the snes emulators bsnes-libretro and bsnes-mercury. Running make in gdb makes its seem like the backtrace is in glibc, but it’d be better with debugging symbols.
https://github.com/libretro/bsnes-mercury/issues/47
I can build it myself if need be, is there anything specific I need to know to accomplish this?
In that case you will have to compile it yourself, as I do not have the package. Obtain Slackware’s glibc source directory, copy the glibc-multilib.SlackBuild script into that source directory, and execute it. A long while later, you will have the glibc packages in a subdirectory “/glibc-/” in the root of your local filesystem.
Hi – heads-up to a possible problem / request for any ideas on how to fix….probably a Current problem rather than your builds….
The problem is that all versions of Slackware builds of Libreoffice found through pkgs.org including your 5.3.1 build use 100% of 1 cpu core continuously when run in a system built with Slackware-Current components from:
Wed May 10 22:58:28 UTC 2017
on an Nvidia based system.
Only seems to affect LibreOffice – for instance Chromium runs as normal with no excessive cpu useage.
peebee, you could check the status of hardware acceleration for graphics output (Tools > Options > Libreoffice > View) and see if it makes any difference if you disable hardware acceleration.
Hi Eric – many thanks for the rapid reply – sadly that control does not make any difference….The cpu useage does drop to 0% if a sub-window such as About LibreOffice is opened but returns to 100% as soon as the sub-window is closed. Also now had chance to check on a laptop with Intel graphics and that has same problem so its not nvidia related. Cheers Peter.
peebee, I tried on my own laptop with Intel graphics, and Slackware64-current with Plasma 5 and I could not reproduce your issue. What program exactly do you have active (writer, calc, …) and what file do you have open? Also, what are you doing with the program?
I can confirm the same LibreOffice problem on Slackware64-current.
You don’t have to be doing anything, just opening LibreOffice and letting it sit idle and one core runs at 99%.
I also see the LibreOffice problem on Slackware64-current. Running “strace -ttt -s 256 -f -p `pidof soffice.bin` -o /tmp/slog” shows repeating polling messages.
— snip —
1866 1494734991.074006 poll([{fd=8, events=POLLIN}, {fd=13, events=POLLIN}], 2, 1009) = 1 ([{fd=8, revents=POLLIN}])
1866 1494734991.074051 read(8, “\1\0\0\0\0\0\0\0”, 16) = 8
1866 1494734991.074093 write(8, “\1\0\0\0\0\0\0\0”, 8) = 8
1866 1494734991.074133 recvmsg(13, 0x7ffed17c7e30, 0) = -1 EAGAIN (Resource temporarily unavailable)
1866 1494734991.074171 poll([{fd=8, events=POLLIN}, {fd=13, events=POLLIN}], 2, 1009) = 1 ([{fd=8, revents=POLLIN}])
1866 1494734991.074215 read(8, “\1\0\0\0\0\0\0\0”, 16) = 8
1866 1494734991.074257 write(8, “\1\0\0\0\0\0\0\0”, 8) = 8
1866 1494734991.074297 recvmsg(13, 0x7ffed17c7e30, 0) = -1 EAGAIN (Resource temporarily unavailable)
— snip —
Perhaps a recompile against the new gcc is indicated.
@cwizardone and @allend – thank you for confirming – apart from that I’m running 32-bit, I can confirm @cwizardone’s description of behaviour – 100% of 1 cpu core occurs on opening LO – sitting idle with no file open….if you open a sub-window, cpu useage drops to normal value but reverts to 100% when sub-window is closed.
Suggestion for correction/update to 2nd paragraph in http://alien.slackbook.org/dokuwiki/doku.php?id=slackware:network
Remove the obsolete ref to slackbasics
or replace with link to https://rlworkman.net/howtos/slackbasics.pdf
I note the same issue on my machine, slachware 64 -current, updated to the newest kernel and refreshed multilib. Libreoffice seems to actually work ok but does indeed use 100% of one of the 8 threads as revealed by htop. Also if i open a sub-window then cpu use drops to 4% or less, then pops back to 100% when sub-window is closed.
I suspect a LO bug not a Bob problem.
Version: 5.3.1.2
Build ID: SlackBuild for 5.3.1 by Eric Hameleers
CPU Threads: 8; OS Version: Linux 4.9; UI Render: default; VCL: gtk2; Layout Engine: new;
Locale: en-US (en_US.UTF-8); Calc: CL
If this was a LO bug, or if it was related to the library changes in slackware-current, then you should try my libreoffice-5.3.3 packages for slackware-current which are being uploaded at the moment.
Just installed your latest LibreOffice package (Thank you), but, unfortunately, the problem remains.
Thanks Eric…..
Tried: libreoffice-5.3.3-i486-1alien.txz …..
On a system built on 7 May – no problem…
On a system built on 15 May – cpu hogging problem
Which seems to indicate that one of the Slackware Current changes on:
Tue May 9 23:33:37 UTC 2017 or
Wed May 10 22:58:28 UTC 2017
is the culprit. But which one??
I’ll try reverting some of the changes to see if the problem goes away – but any guesses as to likely candidates appreciated 😉
Well, mesa was part of the upgrades on May 10. If you have a Nvidia card and are using the proprietary driver, you will have to re-install that Nvidia proprietary driver because Mesa libs will have overwritten some of the Nvidia ones..
i always uninstall and then reinstall the Nvidia driver before and after each kernel and mesa change, so mesa would not appear to be the problem.
fixed.
removing LO 5.3.1 and installing the libreoffice-5.3.3 packages for slackware-current, and a re-boot fixed the cpu-hogging problem for me. I always reinstall the nvidia drivers after each kernel update. LO seems to run slightly faster.
thanks.
Well…. this is interesting. I checked LibreOffice, again, while in Xfce, which is the DE I’ve been using the last week or two, with the same result, that is, it runs one core at 99%.
Just for chuckles, I logged out, changed the DE to KDE-4,
and LibreOffice runs as it should, that is, at idle it uses 4 to 5% of a core.
Whatever that means, I don’t know.
🙂
I’ve tied the culprit down to the /usr/lib libraries in glib2-2.52.2-i586-1.txz (32-bit)
If I revert to the libraries from glib2-2.52.1-i586-1.txz then the problem goes away and cpu useage in libreoffice-5.3.3-i486-1alien.txz returns to normal.
I’ll try to track down exactly which library from glib2 is causing the problem…..
By a process of elimination….
Reverting just: /usr/lib/libglib-2.0.so.0.5200.2
back to:
/usr/lib/libglib-2.0.so.0.5200.1
cures the problem
I’ve been in contact with Robby Workman and he has identified the glib2 update required.
I am curious peebee how you determined that the glib2 libraries were causing this… did you run a trace somehow?
I wish lol – nothing so sophisticated I’m afraid….
just a process of trial and error – overlaid the complete malfunctioning glib2 with the previous version then gradually removed items from the overlay until the problem reappeared….
Does your latest VLC package, #2.2.5.1, address the security issue in this morning’s news?
Quote:
Millions of Kodi, VLC and Popcorn Time users at risk from hackers – update NOW
0
By Jonathan Fairfield on 2017-05-25 Apps & Software, Security
A A A
Users of Kodi, VLC and Popcorn Time are being urged to download a new security update following the discovery of a vulnerability that could allow hackers to take control of your device.
According to security firm Check Point who discovered the vulnerability, hackers are attacking users via the subtitle function, with millions of users thought to be at risk.
The cyber attack takes place when subtitles are loaded onto the video streaming service and has been allowed to go unnoticed as antivirus software typically does not flag subtitles as a threat.
Checkpoint estimates that 200 million users are currently vulnerable to attack.
“By crafting malicious subtitle files, which are then downloaded by a victim’s media player, attackers can take complete control over any type of device via vulnerabilities found in many popular streaming platforms, including VLC, Kodi (XBMC), Popcorn-Time and stream.io”, Checkpoint said in a blog post.
“We estimate there are approximately 200 million video players and streamers that currently run the vulnerable software, making this one of the most widespread, easily accessed and zero-resistance vulnerability reported in recent years.”
Kodi hack
What makes the vulnerability even more dangerous is that it requires no action from the user in order to take place.
There’s no clicking on a link, for example, or being tricked into opening a malicious email or document.
“This method requires little or no deliberate action on the part of the user, making it all the more dangerous,” Checkpoint said.
Once the attack takes place, hackers are able to take control of the device running the streaming service, be it a computer, smart TV, phone, Android box or tablet.
“The potential damage the attacker can inflict is endless, ranging anywhere from stealing sensitive information, installing ransomware, mass Denial of Service attacks, and much more,” Checkpoint said.
Kodi meanwhile have released a statement to say they are aware of the issue and are urging all users to upgrade to the latest version.
“Our developers fixed this security gap and have added the fix to this v17.2 release. As such we highly encourage all users to install this latest version!
“Any previous Kodi version will not get any security patch. We have began the roll out of this version and Android Play Store as well as Windows Store have this update pending and will roll out as soon as possible,” the firm said.
Unquote.
http://tech.thaivisa.com/millions-of-kodi-vlc-and-popcorn-time-users-at-risk-from-hackers-update-now/21877/?utm_source=newsletter-20170525-1353&utm_medium=email&utm_campaign=news
Yes, fixed in 2.2.5.1: http://blog.checkpoint.com/2017/05/23/hacked-in-translation/ see section “Platforms Update”.
alienbob wrote:
“Yes, fixed in 2.2.5.1: http://blog.checkpoint.com/2017/05/23/hacked-in-translation/ see section “Platforms Update”.”
Many thanks!
🙂
Please Add a .tgz in slackbuild.org the programs:
curlew 0.2.4
quelea.org
and openlp 2.4.6
Marlon, I am not the same as slackbuilds.org.
If you want to see new scripts on that web site, you either write them yourselves and submit them for review, or you ask on the slackbuilds-users mailing list whether someone is willing to create them for you.
And in case you wanted to ask again – I am not interested, I do not even know any of these programs so I am unlikely to ever be using them.
Hrm… Perhaps I’m doing something wrong, but after applying a slackpkg update and allowing it to do the “whole thing” including the gcc and libc upgrades, I’m no longer able to build many libraries (like OpenSSL) with pie and hardening options like I used to.
The same build scripts work fine under Gentoo/Ubuntu, incidentally. (I use them to build statically linked OSSL versions of nginx, etc.)
I get nasty errors during the linking phase: /usr/lib64/gcc/x86_64-slackware-linux/5.4.0/../../../../x86_64-slackware-linux/bin/ld: libcrypto.a(cryptlib.o): relocation R_X86_64_PC32 against symbol `stderr@@GLIBC_2.2.5′ can not be used when making a shared object; recompile with -fPIC
/usr/lib64/gcc/x86_64-slackware-linux/5.4.0/../../../../x86_64-slackware-linux/bin/ld: final link failed: Bad value
The command that generates this problem:
gcc -fPIC -fPIE -DOPENSSL_PIC -DOPENSSL_THREADS -D_REENTRANT -DDSO_DLFCN -DHAVE_DLFCN_H -Wa,–noexecstack -m64 -DL_ENDIAN -fPIC -D_FORTIFY_SOURCE=2 -fPIE -Wl,-z,now -Wl,-z,relro -pie -fstack-protector-strong -O3 -Wall -DOPENSSL_IA32_SSE2 -DOPENSSL_BN_ASM_MONT -DOPENSSL_BN_ASM_MONT5 -DOPENSSL_BN_ASM_GF2m -DRC4_ASM -DSHA1_ASM -DSHA256_ASM -DSHA512_ASM -DMD5_ASM -DAES_ASM -DVPAES_ASM -DBSAES_ASM -DWHIRLPOOL_ASM -DGHASH_ASM -DECP_NISTZ256_ASM -Iinclude \
-DFINGERPRINT_PREMAIN_DSO_LOAD -o fips_premain_dso \
fips_premain.c fipscanister.o \
libcrypto.a -ldl
MOST of the time, I am able to add:
-fPIC -D_FORTIFY_SOURCE=2 -fPIE -Wl,-z,now -Wl,-z,relro -pie -fstack-protector-strong
I use grsecurity kernels, and have been aiming to get as much of userspace built with -pie and -z relro/-z now as possible.
With slackware64-current as of about a year ago, I had no problems.
Robin Sryala, the gcc-5.4.0 which you seem to have installed is not a version that’s in any Slackware release. Slackware 14.2 has 5.3.0 while slackware-current moved on to 7.1.0.
Better check the way you upgraded, because if you are still running slackware-current you have ended with the wrong packages.
Hi Eric,
First thanks for your hard & awesome work for this wonderful distro!
I’m using Slackware stable 14.2 with your ktown, multilib, alienbob & restricted repos (up-to date system, as I’m writing this post) but it seems my digikam installation is broken, as indicated by this error when launching the app:
digikam: error while loading shared libraries: libmarblewidget-qt5.so.26: cannot open shared object file: No such file or directory
The library in question is included in your marble-17.04.1-x86_64-1alien package, but it’s not the same version: libmarblewidget-qt5.so.0.27.1 instead of .so.26
Creating a symlink did the trick (ln -s /usr/lib64/libmarblewidget-qt5.so.0.27.1 /usr/lib64/libmarblewidget-qt5.so.26), but shouldn’t digikam be recompiled with the newest library?
Thanks and cheers from Belgium,
K.
Oh! Let me check my command history to confirm this, but I have a suspicion that I may have rolled back gcc (manually with remove/installpkg) after seeing this problem crop. To avoid wasting your time, which is dearly appreciated mind you, I’ll double-check everything again with 7.1.0.
That “bad value” during link almost always comes from a non-PIC object being present in one or more link object(s). Are there intentions to make as much of the userspace binaries -pie/-z relro,now as possible?
Oh, it’s definitely broken when I try to build openssl-1.0.2k with even just -fPIE.
make[2]: Entering directory ‘/dev/shm/tmp/openssl-1.0.2k’
[ -z “” ] || gcc -fPIC -fPIE -DOPENSSL_PIC -DOPENSSL_THREADS -D_REENTRANT -DDSO_DLFCN -DHAVE_DLFCN_H -Wa,–noexecstack -m64 -DL_ENDIAN -fPIC -D_FORTIFY_SOURCE=2 -fPIE -fstack-protector-strong -O3 -Wall -DOPENSSL_IA32_SSE2 -DOPENSSL_BN_ASM_MONT -DOPENSSL_BN_ASM_MONT5 -DOPENSSL_BN_ASM_GF2m -DRC4_ASM -DSHA1_ASM -DSHA256_ASM -DSHA512_ASM -DMD5_ASM -DAES_ASM -DVPAES_ASM -DBSAES_ASM -DWHIRLPOOL_ASM -DGHASH_ASM -DECP_NISTZ256_ASM -Iinclude \
-DFINGERPRINT_PREMAIN_DSO_LOAD -o fips_premain_dso \
fips_premain.c fipscanister.o \
libcrypto.a -ldl
make[3]: Entering directory ‘/dev/shm/tmp/openssl-1.0.2k’
make[4]: Entering directory ‘/dev/shm/tmp/openssl-1.0.2k’
/usr/lib64/gcc/x86_64-slackware-linux/7.1.0/../../../../x86_64-slackware-linux/bin/ld: libcrypto.a(cryptlib.o): relocation R_X86_64_PC32 against symbol `stderr@@GLIBC_2.2.5′ can not be used when making a shared object; recompile with -fPIC
/usr/lib64/gcc/x86_64-slackware-linux/7.1.0/../../../../x86_64-slackware-linux/bin/ld: final link failed: Bad value
collect2: error: ld returned 1 exit status
make[4]: *** [Makefile.shared:169: link_a.gnu] Error 1
The make TABLE output of the linux-x86_64 build settings I’m using:
*** linux-x86_64
$cc = gcc
$cflags = -m64 -DL_ENDIAN -fPIC -D_FORTIFY_SOURCE=2 -fPIE -fstack-protector-strong -O3 -Wall
$unistd =
$thread_cflag = -D_REENTRANT
$sys_id =
$lflags = -ldl
$bn_ops = SIXTY_FOUR_BIT_LONG RC4_CHUNK DES_INT DES_UNROLL
$cpuid_obj = x86_64cpuid.o
$bn_obj = x86_64-gcc.o x86_64-mont.o x86_64-mont5.o x86_64-gf2m.o rsaz_exp.o rsaz-x86_64.o rsaz-avx2.o
$ec_obj = ecp_nistz256.o ecp_nistz256-x86_64.o
$des_obj =
$aes_obj = aes-x86_64.o vpaes-x86_64.o bsaes-x86_64.o aesni-x86_64.o aesni-sha1-x86_64.o aesni-sha256-x86_64.o aesni-mb-x86
_64.o
$bf_obj =
$md5_obj = md5-x86_64.o
$sha1_obj = sha1-x86_64.o sha256-x86_64.o sha512-x86_64.o sha1-mb-x86_64.o sha256-mb-x86_64.o
$cast_obj =
$rc4_obj = rc4-x86_64.o rc4-md5-x86_64.o
$rmd160_obj =
$rc5_obj =
$wp_obj = wp-x86_64.o
$cmll_obj = cmll-x86_64.o cmll_misc.o
$modes_obj = ghash-x86_64.o aesni-gcm-x86_64.o
$engines_obj =
$perlasm_scheme = elf
$dso_scheme = dlfcn
$shared_target= linux-shared
$shared_cflag = -fPIC -fPIE
$shared_ldflag = -m64
$shared_extension = .so.$(SHLIB_MAJOR).$(SHLIB_MINOR)
$ranlib =
$arflags =
$multilib = 64
Random slacker, indeed digikam needs to be recompiled.
Robin, I think you are attempting something that is out of my scope and out of scope for the Slackware distribution.
If you get the same issues compiling software with PIE when you use the default Slackware gcc packages instead of my multilib packages, then there’s nothing I can change about it. If Slackware’s own gcc enables you to do what you want but the multilib version of gcc cannot, then it is an issue with my multilib packages and I can try to spend some time looking into it, but really: Slackware is not going the road of hardening its binary code. In any case , this is not something you should ask me. It’s only Patrick Volkerding who can answer that question.
It seems like the error is suggesting that some glibc library that should be built with PIC hasn’t actually been done so.
However, I just tried the same build script on Ubuntu Zesty, and get the same errors, so perhaps something’s changed within OpenSSL between 1.0.2g and 1.0.2k to completely break PIE builds. (The problem is always the fipscanister object.)
Ah, thanks for the clarification, Eric. So Patrick is still active within the Slackware project? It was my first distro all of that time ago in the 1990s, and I still like the old-style “nothing installed except for what I put there” ethos of it. While I have tried most others, I continue to get annoyed by things changing config files in /etc or elsewhere “behind my back”. Slackware remains a kind of “soft warm blanket” of Linux distros for me.
Thank you (and Patrick) for continuing to keep the project going.
Robin: Slackware is Patrick Volkerding. I am just a concerned 3rd party, or else part of a coreteam of volunteers helping Pat out with the research & testing which precedes public distro updates.
Understood, Eric. My thanks for your maintaining this resource and being so responsive, even in the face of noise like I’ve generated here.
I’ll take up the issue (in general) of aiming to harden the userspace binaries of Slackware. While it’s easy to replace the kernel, it’s much harder and more time consuming to rebuild + replace all of userspace to make the most of it.
I can appreciate that supporting foreign (often non-GPL derived) binaries in a fully functional ecosystem is only made that much harder when dealing with “extra features” like those.
Thanks, Eric, for all your work.
I’m using your wine-2.8-x86_64-1alien package on Slackware 14.2 multilib. When I run wine –check-libs, most of the libraries are found, but a few aren’t. Specifically, it says that gnutls is missing:
libgnutls.so.30: missing
And, when I try to run a Windows program, I see the following error:
err:winediag:schan_imp_init Failed to load libgnutls, secure connections will not be available.
However, I see it on my system:
lrwxrwxrwx 1 root root 20 May 3 19:31 /usr/lib/libgnutls.so.30 -> libgnutls.so.30.13.1
-rwxr-xr-x 1 root root 1537368 Jan 10 20:16 /usr/lib/libgnutls.so.30.13.1
lrwxrwxrwx 1 root root 20 Jan 18 13:11 /usr/lib64/libgnutls.so.30 -> libgnutls.so.30.13.1
-rwxr-xr-x 1 root root 1484608 Jan 10 20:30 /usr/lib64/libgnutls.so.30.13.1
$ ls /var/log/packages/ | grep gnutls
gnutls-3.5.8-x86_64-1_slack14.2
gnutls-compat32-3.5.8-x86_64-1_slack14.2compat32
Any thoughts on what the issue is here? Thanks.
Ben, my guess is that you do not have a fully up-to-date multilib installation. From time to time, I add new packages to the “compat32” set when other packages get new dependencies.
Run the command “ldd /usr/lib/libgnutls.so” to see if there are any missing libraries (the “not found” errors). It’s probably libunistring which you are missing.
If you are using the slackpkg+ extension to slackpkg, then the solution is to simply run “slackpkg install multilib” which will install all the missing packages. Otherwise, you need to download all packages below http://www.slackware.com/~alien/multilib/14.2/slackware64-compat32/ and run “upgradepkg –install-new” on them:
Hi Eric,
just a remark on the vpnc script you have in your openconnect package. It erroneously calls netconfig, thinking we are on SuSE. I just commented out the line pointing to suse…
diff vpnc-script vpnc-script.orig
130,132c130,132
< elif [ -x /sbin/netconfig ]; then # tool on Suse after 11.1
< MODIFYRESOLVCONF=modify_resolvconf_suse_netconfig
#elif [ -x /sbin/netconfig ]; then # tool on Suse after 11.1 commented out FMS
> # MODIFYRESOLVCONF=modify_resolvconf_suse_netconfig
> # RESTORERESOLVCONF=restore_resolvconf_suse_netconfig
A better way to do it is proposed by ponce in linuxquestions:
http://www.linuxquestions.org/questions/slackware-14/vpnc-on-slackware-14-2-is-bringing-up-network-configuration-dialog-each-time-4175595447/#post5646866
Only he says line 128, and I think it is 130
Anyway, thanks for your wonderful work. You’ve saved me and others tons of hours and frustration….
Good luck
Franz
Thanks so much, Eric! That was exactly the issue.
Franz, I will apply ponce’s patch and release an updated package.
I do not use openconnect myself, and added it only because someone wanted that I compile plasma5-nm against it.
Thank you! I’m also _very_ happy not to have to include the stuff myself 🙂
Have a nice time
Franz
Today’s multilib updates seem to have fouled up my WINE installation. When I try and run a program that ran fine before the updates, the following error results.
**********************************************
fixme:winediag:start_process Wine Staging 2.8 is a testing version containing experimental patches.
fixme:winediag:start_process Please mention your exact version when filing bug reports on winehq.org.
Wine cannot find the FreeType font library. To enable Wine to
use TrueType fonts please install a version of FreeType greater than
or equal to 2.0.5.
http://www.freetype.org
wine: Unhandled page fault on execute access to 0x00000000 at address (nil) (thread 0038), starting debugger…
Wine cannot find the FreeType font library. To enable Wine to
use TrueType fonts please install a version of FreeType greater than
or equal to 2.0.5.
http://www.freetype.org
err:syslink:SYSLINK_SetFont Failed to create link font!
*****************************************************************
Just FYI.
🙂
cwizardone check your 32bit libraries with ‘ldd’. Have you done “slackpkg install multilib” recentluy to get any newly added compat32 packages?
Turns out the problem was the latest NVIDIA “long lived branch” beta driver 384.47.
I reverted back to 375.66 and the Wine problem disappeared.
I did however, download fresh copies of all the multilib files. Only one, the alsa-plugins, was out of date.
Hei Erik, I’ve a question for you.
Do you know why your live ISO for 14.2 ( and also the original one ) does not see my disk partition ?
With both iso I can mount my exisiting partitions, but when I run the setup or the setup2hd I alway get a message saying that there are no partition on disk and I need to run fdisk or gdisk to create one.
Just for know, I have an MacBookPro that is happly running slackware 14.1 x86_64 multilib.
Pierluigi – you need to be more specific in your details. I can not look into your computer so I can not tell you why Slackware setup fails to recognize your disk partitions.
The installer uses a script called “/sbin/probe”, it is part of the installer’s initrd so you will not find it on a regular Slackware OS. If you boot the Live OS, you will find the “probe” script as “/usr/share/liveslak/probe”. Try running this command:
/usr/share/liveslak/probe -l 2>/dev/null | grep -E ‘Linux$’ | sort
… because that exact command is also executed by ‘setup’ to determine the presence of Linux partitions. Tell me what that gives.
If the command output is empty, then the output of “/usr/share/liveslak/probe -l” would be interesting.
… and it is Eric – with a ‘c’. I get tired of people who read my blog, post comments and get my name wrong. How simple it is to be polite and respectful, and write my name correctly.
Eric, please apologize on getting your name wrong. And yes, I should have checked before writing.
I will try to get the output of those command and post the result in a little time.
Here the outputs:
************************************
Output of “gdisk -l /dev/sda”
GPT fdisk (gdisk) version 1.0.1
Partition table scan:
MBR: hybrid
BSD: not present
APM: not present
GPT: present
Found valid GPT with hybrid MBR; using GPT.
Disk /dev/sda: 976773168 sectors, 465.8 GiB
Logical sector size: 512 bytes
Disk identifier (GUID): 5BF422E2-E139-4A5A-A7F3-3A5BC14CD5DE
Partition table holds up to 128 entries
First usable sector is 34, last usable sector is 976773134
Partitions will be aligned on 8-sector boundaries
Total free space is 28999005 sectors (13.8 GiB)
Number Start (sector) End (sector) Size Code Name
1 40 409639 200.0 MiB EF00 EFI System Partition
2 409640 195722135 93.1 GiB AF00 Mavericks
3 195722136 196991671 619.9 MiB AB00 Recovery HD
4 197255168 302112767 50.0 GiB EF00 Linux
5 302112768 406970367 50.0 GiB EF00 Topo
6 406970368 423747583 8.0 GiB 8200 swap
7 423747584 633462783 100.0 GiB 0700 extbig
8 633462784 738320383 50.0 GiB 0700 ext
9 738320384 843180031 50.0 GiB 0700 ext1
10 843180032 948037631 50.0 GiB EF00 ext3
************************************
output of :”/usr/share/liveslak/probe -l 2>/dev/null | grep -E ‘Linux$’ | sort”
************************************
Output of: “/usr/share/liveslak/probe -l ”
Disk /dev/sda: 465.8 GiB, 500107862016 bytes, 976773168 sectors
Units: sectors of 1 * 512 = 512 bytes
Sector size (logical/physical): 512 bytes / 512 bytes
I/O size (minimum/optimal): 512 bytes / 512 bytes
Disklabel type: gpt
Disk identifier: 5BF422E2-E139-4A5A-A7F3-3A5BC14CD5DE
Device Start End Sectors Size Type
/dev/sda1 40 409639 409600 200M EFI System Partition
/dev/sda2 409640 195722135 195312496 93.1G Apple HFS/HFS+
/dev/sda3 195722136 196991671 1269536 619.9M Apple boot
/dev/sda4 197255168 302112767 104857600 50G EFI System Partition
/dev/sda5 302112768 406970367 104857600 50G EFI System Partition
/dev/sda6 406970368 423747583 16777216 8G Linux swap
/dev/sda7 423747584 633462783 209715200 100G Microsoft basic data
/dev/sda8 633462784 738320383 104857600 50G Microsoft basic data
/dev/sda9 738320384 843180031 104859648 50G Microsoft basic data
/dev/sda10 843180032 948037631 104857600 50G EFI System Partition
***************************************
When the 14.1 is running I have:
Output of : ” df -h”
Filesystem Size Used Avail Use% Mounted on
/dev/sda4 50G 38G 9.4G 80% /
/dev/sda1 197M 130M 68M 66% /boot/efi
tmpfs 3.9G 101M 3.8G 3% /dev/shm
/dev/sda7 99G 89G 5.1G 95% /extbig
/dev/sda8 50G 31G 17G 65% /ext
/dev/sda9 50G 27G 21G 56% /ext1
/dev/sda5 50G 37G 11G 79% /mnt/sdb2
Output of : sudo mount
sudo mount
/dev/sda4 on / type ext4 (rw)
proc on /proc type proc (rw)
sysfs on /sys type sysfs (rw)
/dev/sda1 on /boot/efi type vfat (rw)
tmpfs on /dev/shm type tmpfs (rw)
/dev/sda7 on /extbig type ext3 (rw,nosuid,nodev,noatime,nodiratime,errors=remount-ro)
/dev/sda8 on /ext type ext3 (rw,nosuid,nodev,noatime,nodiratime,errors=remount-ro)
/dev/sda9 on /ext1 type ext3 (rw,nosuid,nodev,noatime,nodiratime,errors=remount-ro)
gvfsd-fuse on /home/pigi/.gvfs type fuse.gvfsd-fuse (rw,nosuid,nodev,user=pigi)
/dev/sda5 on /mnt/sdb2 type ext3 (rw)
Output of: gdisk -l /dev/sda
GPT fdisk (gdisk) version 0.8.7
Partition table scan:
MBR: hybrid
BSD: not present
APM: not present
GPT: present
Found valid GPT with hybrid MBR; using GPT.
Disk /dev/sda: 976773168 sectors, 465.8 GiB
Logical sector size: 512 bytes
Disk identifier (GUID): 5BF422E2-E139-4A5A-A7F3-3A5BC14CD5DE
Partition table holds up to 128 entries
First usable sector is 34, last usable sector is 976773134
Partitions will be aligned on 8-sector boundaries
Total free space is 28999005 sectors (13.8 GiB)
Number Start (sector) End (sector) Size Code Name
1 40 409639 200.0 MiB EF00 EFI System Partition
2 409640 195722135 93.1 GiB AF00 Mavericks
3 195722136 196991671 619.9 MiB AB00 Recovery HD
4 197255168 302112767 50.0 GiB EF00 Linux
5 302112768 406970367 50.0 GiB EF00 Topo
6 406970368 423747583 8.0 GiB 8200 swap
7 423747584 633462783 100.0 GiB 0700 extbig
8 633462784 738320383 50.0 GiB 0700 ext
9 738320384 843180031 50.0 GiB 0700 ext1
10 843180032 948037631 50.0 GiB EF00 ext3
So both gdisk seems to have the same output. And ( as far as I have read ) I have the EF00 partition[s] on GPT for installer to work.
Btw I would like to install it on /dev/sda10 or at least on /dev/sda5.
And I can’t even understand why the gdisk report the name on last column with those names. I haven’t named thems ( nor as I can remember )
Thanks for your time, and apologize again.
Eric, I did some other tests: in the “probe” script it seems that the “Linux$” partition is found only if there is a partition type 8300 in the GPT (around line 238 ).
I’ve changed the type from ef00 to 8300 and now the setup2hd can see a partition where to install.
I wonder if this is right, as also ef00 is well working ( and sometimes advised ) for the ext3/ext4 filesystem ( at least so I’ve googled around ) .
Hi Pierluigi,
Indeed the Slackware setup only looks for Linux partitions (gdisk type 8300). On a computer that adheres to the UEFI standard, the EFI System Partition (ESP) should be a single partition, formatted with a filesystem compatible with FAT (https://en.wikipedia.org/wiki/EFI_system_partition). This partition is read by the UEFI firmware and the various OS-es are loaded from it.
You appear to have multiple partitions labeled as ESP and then you formatted two of them with an ext4 filesystem instead. This is not according to UEFI specs.
Once you changed the partition type to 8300 everything was in order again, and Slackware setup had no more problem.
PS: The partition text labels shown by gdisk are written into the partitions by programs like gdisk and others. I do the same in my iso2usb.sh script which is part of liveslak. It is helpful to people who are not so familiar with partition types.
Hi Eric,
thank a lot for your time !
And please, apologize again.
Hi Eric!
I highly appreciate your work on Slackware and, I think, I found a bug in your Chromium package – it simply not playing ogg video format (.ogv).
This is prooflink for that:
https://matrix.org/_matrix/media/v1/download/matrix.org/xDfUNHDLRcMgHcRHunCBJQhg
Thanks a lot for your time and attention!
No idea what that URL is supposed to play, but according to this page, Ogg/Theora and WebM video playback works: https://www.quirksmode.org/html5/tests/video.html#link4
Here is another Ogg video that plays properly in my Chromium browser: http://techslides.com/demos/sample-videos/small.ogv
Indeed, your samples is working good, and Ogg/Therora playback as well. But (https://matrix.org/_matrix/media/v1/download/matrix.org/xDfUNHDLRcMgHcRHunCBJQhg) is not, though it works on Firefox 54/Windows 8.1. Maybe this is somewhat proprietary?
Garik I tried that URL with Chrome on Windows and it does not play there either. Therefore it is not a bug in my package. I downloaded the file behind that URL and it is a real Ogg Video which plays fine with ffplay for instance.
Feel free to open a bug report in Chromium’s bugtracker.
Thanks for clarification, Eric. It’s clearly Chromium upstream bug, since all other .ogv are working, the problem is only with videos without .ogv extension in the address bar.
will you be going over to the newer ffmpeg and remove the faac now that aac is actually part of ffmpeg edited your script to build with libfdk-aac because the newer ffmpeg –enable-faac is no longer used same as faad also.
or are you going to let the team handle it now that it is part of current and lamemp3 is now open?
Currently running VLC 3.0 git and that was a ton of work.
with your system still working on it.
In my ffmpeg.SlackBuild and the package, I removed faac/faad2 support in the 3.2.4 release (Feb 2017). But I think you were not referencing that.).
When updating the VLC package I always follow the developer’s choice of ffmpeg sources, so I never use the latest release and take a specific snapshot which is validated by the VLC developers. The ffmpeg source tarball I use for vlc-2.2.6 actually still supports “enbable-libfaac”.
The faad2 library is not used at all in the internal ffmpeg library for VLC. It’s part of the configure command for VLC itself ( –enable-faad –with-faad-tree=
Hi eric,
Your nodejs package for slackware64-current contains a file that has not been properly sedded and thus contain a wrong path (lib instead of lib64).
The file is /usr/lib64/node_modules/npm/lib/npm.js. It contains at line 353 :
? path.resolve(npm.globalPrefix, ‘lib’, ‘node_modules’)
Which seems to change the install behavior of npm. For me at least.
I change “lib” for “lib64” and npm installed packages to /usr/lib64 after instead of /use/lib.
gunbird – I will look into that. Thanks for reporting.
Hello Linux newbie here. I tried to google answers but failed). I have customized /home which I copied manually from the running Slackware Live Edition file system. I want Slack Live Ed to mount it as /home next time I boot it. I thought about something like syslinux boot parameters – I tried a few combinations with luksvol but my /home is a directory not an encrypted image so luksvol=home:/home didn’t work. I dont have any persistence specified. I am using your newest 700~mb iso xfce64bit…from either virtualized or real cdrom.
zeedr – what you try to achieve is currently not supported in Slackware Live Edition.
It is not hard to add this functionality though, so I might add it to my TODO list for liveslak.
Thanks Eric for the quick reply. Btw, it is a great idea with the Live Edition as a demonstration of a current.
Hi mr. Hameleers. is there the possibility of upgrading your electrum bitcoin wallet package to the latest release? And many thanks for the new plasma 5 packages, working great so far.
Hi fabio,
Yes I will add that to the TODO list, my package is pretty old already.
Hi Eric. Just wondering how the new machine is shaping up?
It’s running just fine, but I currently do not have the time to configure it for my needs. Real life is taking priority at the moment.
I am experiencing some strange screen behavior with the most recent libreoffice
[Version: 5.3.4.2
Build ID: SlackBuild for 5.3.4 by Eric Hameleers
CPU Threads: 8; OS Version: Linux 4.9; UI Render: default; VCL: kde4; Layout Engine: new;
Locale: en-US (en_US.UTF-8); Calc: CL]
I am running slackware64 -current -multilib and have a i7-7700k & kabylake motherboard and nvidiia 760 graphics card.
I am getting overwriting of numbers in the menus (selected values overwrite themselves rather than replace themselves so that i get a jumble of overwritten numbers.
occasionally i see screen tearing but no hanging of the machine. No problems with any other software.
I have tried reloading nvidia driver but didnt help.
any advice?
Wayne
Wayne, are you using a non-default theme in KDE4? If so, try using the KDE4 default. Could also be a graphics driver bug, the LibreOffice bugtracker is full of graphical menu glitches. People have been submitting those for years.
I am not yet certain it is a good idea, but I fixed my troubles for now by installing the Open Office 4.1.3 SlackBuild.
It seems to run ok on the same machine as LibreOffice 5.3.x without any communication or interference and it is quite similar.
If you update to run slackware-current, does the /etc/slackware-version file show you are running slackware-current? ir dies it still show the last stable release version (14.2)?
I’m trying to tell if a certain computer is running slackware -current and what date it was last updated to -current.
Thanks
Steve, the version in the file “/etc/slackware-version” will mention 14.2 even when you are running -current. Its version will only change when the next release is being readied.
The best way to determine whether a particular computer was installed with “slackware-current” and at what point in its development (perhaps even between 14.1 and 14.2 releases) you should check the /var/log/packages directory and find the most recently installed official Slackware packages.
Then determine from the Slackware-current ChangeLog.txt (and possibly the ChangeLog.txt of older releases) how stale your -current installation is. If this is a Slackware-current install then the timestamps of the files in /var/log/packages should help you as much as the package version numbers.
Just a heads up.
The updated package, freetype-compat32-2.8.1-x86_64-1compat32, will cause WINE to fail and return the
error,
bash-4.4$ env WINEPREFIX=”/home/mwnn/.wine” wine “C:\AHDW\AHD3.EXE”
fixme:service:scmdatabase_autostart_services Auto-start service L”AFS2K” failed to start: 3
wine: Unhandled page fault on read access to 0x00000004 at address 0x7eb5277b (thread 0035), starting debugger…
winedbg: Internal crash at 0xf6f7d77b
bash-4.4$
Reverting to the previous version solves the problem. Apparently, there has been some problem with the same
package for -current and Firefox.
cwizardone I’ll update the multilib freetype-compat32 package soon.
I have to respin the PLASMA5 Live ISO image too because the one I just generated earlier today contains that same multilib package. I was uploading it when I saw your post.
Hello Eric. Thank You for providing Palemoon packages, and slackbuild. Really liking this browser as compared to the growing firefox. I’m having issues building a package myself though; even without any modifications to the slackbuild. I’m on the latest current x64. Are there any extra dependencies that are required?
Boris in any case you need to compile using my gcc5 package; the gcc7 in slackware-current will not be able to compile palemoon.
Boris I have uploaded Pale Moon 27.5.0 packages, with a single edit to the SlackBuild.
Did not know gcc5 was required. I was just able to successfully build the package myself. Thank You.
There is a bug in your Slackbuild concerning the optimizations. So if I change line 85 from 0 to 1, line 267 unintentionally expands the variable to multiple lines.
ac_add_options –enable-optimize=”-O2
ac_add_options -msse2
ac_add_options -mfpmath=sse”
Preventing a correct .mozconfig file, and causing a build error.
I removed line 267 and added this line directly after the options for loop; as I wanted to have an optimized build.
echo “ac_add_options –enable-optimize=\”-O2 -msse2 -mfpmath=sse\”” >> .mozconfig
Boris, good catch. I have applied your fix to the palemoon.SlackBuild, thanks.
Dear AlienBOB – you make a real good job. People like you are rare.
Hi, Alien i’m running Slackware 32 bit 14.2
I wonder if i can put in a package request for phpxmlrpc
I’m trying to install moodle at /var/www/htdocs/moodle & output is:
php_extension xmlrpc needed.
Got feedback from willysr on LQ.org who said “hat’s included in Slackware Current, but not yet in stable” so will have to build yourself
Found source at: https://github.com/gggeek/phpxmlrpc/releases/tag/4.2.0
Having a go with src2pkg (with -VV) but its hanging at: ” Checking support for DESTDIR (or similar) – ”
Will have a go with build script if theres a very low learning curve guide
andy, what willysr meant is that the PHP package in slackware-current was extended with the rpc extension. You are looking at the wrong source.
What you can try is take the Slackware source for php package from slackware-current tree: http://bear.alienbase.nl/mirrors/slackware/slackware64-current/source/n/php/
Then, run the php.SlackBuild script and upgrade your php to this new package.
If on the other hand you do not want to install a newer php version you can take the php sources from the slackware-14.2 tree and add a single line to the configure command inside the php.SlackBuild:
–with-xmlrpc=shared \
And then execute that script to recompile a php package *with* a xmlrpc extension included.
ok cheers Alien one question i’m on 32 bit is http://bear.alienbase.nl/mirrors/slackware/slackware64-current/source/n/php/
for 64 bit slackware only?
andy, the package sources are independent of the archtecture, their source trees are completely identical.
You can download and use the sources from a 64bit Slackware tree on your 32bit OS.
well went for –with-xmlrpc=shared \
I got build script from : http://slackware.uk/salix/i486/slackware-14.2/source/n/php/
I put –with-xmlrpc=shared \ in section
CFLAGS=”$SLKCFLAGS” \
./configure \
–prefix=/usr \
ran & it cam e up with line 79 alpine.SlackBuild missing
Downloaded alpine from http://slackware.uk/salix/i486/slackware-14.2/source/n/alpine/
good news I got :
bash-4.3$ php -v
PHP 5.6.23 (cli) (built: Oct 15 2017 13:09:38)
Copyright (c) 1997-2016 The PHP Group
Zend Engine v2.6.0, Copyright (c) 1998-2016 Zend Technologies
with Zend OPcache v7.0.6-dev, Copyright (c) 1999-2016, by Zend Technologies
So i didn’t wreck my system- moodle still moaning it needs xmlrpc
think i’ll give moodle a miss while my slackware is strll stable!
andy, I prefer it if you stop duplicating your posts on LQ.org here on my feedback page. Your questions have nothing to do with my stuff. Leave it at LQ please.
Hi Eric, I downloaded your kdenlive-17.08.2-x86_64-1alien.txz package and related deps. Post install kdenlive won’t start exiting with “kdenlive: error while loading shared libraries: libKF5KIOFileWidgets.so.5: cannot open shared object file: No such file or directory”. I googled for this lib but no luck. Any tips appreciated!
Thomas, you can not download just kdenlive (and deps… what deps?) and hope it works. You will probably not get it to work unless you install my full Plasma5.
Hi Eric,
dhcpcd entry in rc.inet1
I have been trying slackware64-live current. Out of the box rc.inet1 did not connect to my Epson All-In-One; because it is an IPv4 only device.
After some digging, I dropped the -L option from this line in rc.inet1 (it’s all one line):
/sbin/dhcpcd -L -t ${DHCP_TIMEOUT[$i]:-15} ${DHCP_OPTIONS} ${1}
Would you consider dropping -L as the default from now on?.
Thanks, Thoralf
Thoralf the “-L” option does not have do anything with connecting to a ipv4-only device. The dhcpcd works with ipv4 only, anyway.
What the “-L” does, is thatit will prevent the DHCP client from configuring itself with a Link-Local address (range 169.254.0.0/16) in case it does not find a DHCP server.
Removing the “-L” means that your computer will fail to connect to the network if the DHCP server is too slow in answering, and that happened a lot – which is the reason the “-L” is used.
Your case is an exception, apparently you do not have a DHCP server in your LAN?
Hi, Eric.
chromium-pepperflash-plugin-27.0.0.170-x86_64-1alien needs to be removed from http://www.slackware.com/~alien/slackbuilds/chromium-pepperflash-plugin/pkg64/ in all sub-directories. Otherwise, slackpg+ gets confused. Thanks, and enjoy your Sunday.
Richard, thanks, I have taken care of it.
Hi Eric, how you deal when you are testing a new slackbuild, and it fails several minutes or even hours later, is there a way around this?
I mean, for example, there is a package that I was last week trying to compile, but the new GCC 7.0 is giving me a hard time, and it fails at 76% or above that, do you have a trick or something to start where it fail last time, or there is not and I have to start over and over again?
Pablo, welcome to my world.
It sometimes takes many many restarts of a SlackBuild script and many many hours or even days to make it work. It happens a lot with new chromium, libreoffice, kde, vlc releases for instance.
If you know how a build works you can try finding the root cause of a compilation failure by going into the extracted sources, fix the issue and run “make”. Because “make” works incrementally you don’t have to compile anything that’s already been compiled and you may save time.
On the other hand, if the compilation depends on all kinds of environment variables you have set in the SlackBuild script, it will be quite difficult to work with that outside of the script.
Writing a working SlackBuild is an underrated effort. It takes a lot of time to get complex sourcecode built correctly.
Hi Eric! Back at the beginning of October I posted a comment about an issue I was having with kdenlive from your kde5 installation. What was puzzling was that I had two machines, an old one with amd/ati graphics on which kdenlive ran perfectly, and a new one using intel graphics that kept crashing – always for a different reason and consistent only in that it made the program unusable. No-one came forward with a solution, but I have finally found it!
The problem was that, at some stage in the development of kdenlivet, something in ~/.config/kdenliverc has changed, and using an older version of this file with a newer version of kdenlive made it seriously unstable!
My old machine had never had kdenlive installed, previous to my upgrading to your kde5. The newer machine had previously had an older (kde4 compatible) version prior to the upgrade!
Simply deleting ~/.config/kdenliverc and starting afresh cured all the crashing issues!
Of course, you lose all your default settings doing this, but it doesn’t take long to re-create them.
Sometimes its the simplest things that trip you up! This one certainly caused me a lot of grief until I found the cause!
Many thanks for all your hard work, both on Slackware and kde5!
Cheers,
—
Pete
Peter, thanks for documenting your findings here on my blog, I hope it will be helpful for people who run into this issue later on.
One of the things to try when you run into usabilty issues like these, is to create a fresh new user account, login to that account and see if the failing application behaves differently or even properly with this new account. If it works as it should, then the cause will be buried somewhere in your own user’s configuration files. Mostly, they will be found in ~/.config/
Life is good when some one pats you on the back. KISS. Have beer my friend
Drakeo, thanks for the tip. I will add a good beer to my fridge this weekend.
Thanks for posting that sausage roll recipe again. I’ve been making my own sausage for decades and it looks tasty – especially on these below 0 days.
I would really like to know why you create your own Palemoon build/package? I noticed that palemoon.org links to built packages by khronosschoty; and the Slackbuild exists on Slackbuilds.org. It obviously takes time and effort, so I’m very curious.
Boris – there are two scripts on SlackBuilds.org. The ‘palemoon’ script only re-packages the pre-built binaries from palemoon.org. These binaries were not compiled on Slackware.
The ‘PaleMoon’ script on SBo (written and maintained by khronosschoty) will indeed compilie Pale Moon from source but it requires that you also build and use a ugly (IMHO) gcclegacy494 package which contains a gcc 4.x compiler which will then build the binaries. That PaleMoon script by khronosschoty uses the compiler flags and optimizations as suggested by the Pale Moon developers.
My own palemoon package is built using more conservative optimization parameters, and it is using gcc 5 instead of gcc 4 to compile the source. This is against the developers’ recommendation but I see no need to stick with the old gcc 4.
On Slackware-current I provide a gcc5 package which you can install in parallel with Slackware’s own gcc 7.
Also note, that SBo only provides the build script, no binaries. I provide a binary package with the build script as bonus. You can directly install the package from my repository.
your libtinfo on current is broken. seems pat is building with the compat 5.9 so it goes back to the old link not the libncurses++w.so my google builds work now when going back to ln -s /usr/lib?/libncurses.so.5.9 /usr/lib?/libtinfo.so.5
ok i was wrong but your links are incorrect u
/lib64 or lib need links to libncurses.so.5.9 may want to look at your pkg build script
drakeo there is no need to install libtinfo on Slackware-current anyway. Chromium builds fine these days without it.
my apologies for posting this under the vlc update post. Had two tabs open, and pasted into the wrong one.
todays current upgrade breaks Calibe and Libreoffice due to new icu4c. creating symlinks does not work, and throws an error if they’re created. qt5 was also built with it, so any application that I build requiring qt5 will fail to do so.
to be more specific /usr/lib64/qt5/bin/lrelease is the qt5 file that causes build errors for certain applications due to new icu4c
The qt5 package needs to be recompiled, as do some others. I will slowly get there. Qt5 is a beast.
Think for compat32-tools-3.7-noarch you might want to do this
for us old guys that forget to back up a massconvert32.sh
or something like that for that part of the tarball
config() {
NEW=”$1″
OLD=”$(dirname $NEW)/$(basename $NEW .new)”
# If there’s no config file by that name, mv it over:
if [ ! -r $OLD ]; then
mv $NEW $OLD
elif [ “$(cat $OLD | md5sum)” = “$(cat $NEW | md5sum)” ]; then
# toss the redundant copy
rm $NEW
fi
# Otherwise, we leave the .new copy for the admin to consider…
}
I built a back up into my slackupdate script I run . OH well . I asked pat a long time ago for the link to vmlinuz-generic it worked well. Happy Holidays.
Drakeo, massconvert32.sh is a script, not a configuration file. If you edit a script you should make a copy first. It’s not going to get the .new treatment, sorry.
now testing Linux-4.15-rc3, all good !
I’m adding all possible ‘m’ modules, at amd64 good work.
linux-4.15-rc3-amd.7cyber.txt you can found here:
https://vk.com/7cyber
also screenshot available at facebook.com/roman.dyaba
I’m disable tux logo, just you can add this.
You can add this ti current amd64 mainstream.
Thank you.
About mutex, like a periodic software sheduling
(sample type 00010001 or 01010101 or 1002001003001),
Small and fast kernel, memory management,
see book http://www.swd.ru/print.php3?pid=643.
Also stable XFCE thith KDE5 decoration, see at my posts.
For 32-bit users:
You can use your’s very stable system in future.
Hmmm… Who use 32 bit ?
latest current breaks libreoffice due to boost.
Yes, fun is it not, running -current, Boris? You can rebuild libreoffice if you need it working fast. If you rely on my package to be rebuilt, it can take a while.
Probably more things will show up broken, boost upgrades cause that every time. Just like icu4c and poppler upgrades, those are the ones I hate with a vengeance.
Roman, this blog is not your blog. So stop posting random farts or I will block your future posts.
I personally enjoy running current. Yes every so often these things happen, but it’s ok. In this case libreoffice seems to be the only thing broken by the update. I’ll rebuild it myself I think.
…and no need to build. creating the following symlinks in /usr/lib64 does the job
ln -s libboost_system.so.1.66.0 libboost_system.so.1.65.1
ln -s libboost_iostreams.so.1.66.0 libboost_iostreams.so.1.65.1
ln -s libboost_filesystem.so.1.66.0 libboost_filesystem.so.1.65.1
It’s just that I raised a few acute problems.
Including the Linux kernel. Threads. Memory. Swap/hdd.
I definitely prefer XFCE instead of buggy KDE4/5.
From KDE is only K3Burn & Education/games.
Add good and beautiful games instead of indie games.
And libreoffice.org
Remember qnx-demo-floppy is dedicated.
Thank you.
Many thanks for the quick LibreOffice update. Greatly appreciated.
Merry Christmas to You and Yours!
🙂
Hi. After upgrading kde5 to the December release I’ve got some problems with DigiKam (5.7.0). I cant switch the album in digikam. It shows all albums no meter which album is selected on the left. Does any one have idea where to search for the problems? or does any one have some problems or maybe it only on my system?
Konstantin, a hint. Next time, put your comment below the relevant blog post.
Hello, Eric.
There is a strange problem with last Slackware64-current update. After it I found out that the /opt directory belongs to me. And because of that VirtulBox didn’t work. After chmod root:root /opt it works again.
Best Regards.
Anton; changing /opt ownership to your regular user account is definitely something the Slackware package tools did. You did that. One possibility is that you created a custom package which has an /opt directory inside, with you as the owner. Installing such a package would change the system /opt directory’s ownership to you.
$ ( cd /var/log/packages/; ls | xargs grep ‘^opt’ )
aaa_base-14.2-x86_64-2:opt/
$
https://www.linuxquestions.org/questions/slackware-14/virtualbox-won%27t-start-4175605025/
But who did this?
I’ve just upgraded from 14.2 32 bit to 64 bit, and once again, you have rescued me with two “install resistant” packages I can’t live without; vlc and qemu. Thank you for the hard work, and organisation.
Slackware enthusiast since version 7.
David, good luck with the switch to 64bit. I assume you had to re-install from scratch or did you manage to come up with an in-place upgrade?
I see you have a script to update the kernel on a liveslak USB stick. Can it upgrade the kernel to the testing kernel? I just bought a new laptop with a core i5 8250U and the Slackware64 -current iso boots but keyboard input comes out wrong (even the enter key won’t work). I think it needs a kernel after 4.10.
And great work, by the way.
Joe, it can upgrade to any kernel. The script will accept regular Slackware packages (kernel plus kernel-modules) or else a bare kernel-image file and a directory tree with its modules.
You need to ensure that this kernel contains the support required for booting a Live OS if you want to compile a custom kernel yourself. All Slackware kernels, including those in /testing, should work with liveslak.
Roman, I had warned you. This is not your blog, and you keep on posting random farts. Use linuxquestions.org for that. You are banned here as of now.
Hi Eric
Is there an official way to safely debug slackbuild script that thrash system files ?
Kind regards
Barend
Barend there are many ways to make sure you write a well-behaving SlackBuild script.
If you want to develop one safely, you should consider creating a virtual machine, a LXC or docker container, a LVM snapshot etc which allow you to safeguard against rogue installs and detect files that were installed outside the $PKG directory.
This is a complex issue, and if you want to discuss this in full then perhaps you need to open a topic on linuxquestions.org so that more people can give their feedback.
Muchas Gracias from New Mexico, USA, for convenient Slack live, wine, & multilib packages. I used those along with python.org’s win32 Python27, Kindle for PC 1.7, & Apprentice Alf’s DRM tools to replace broken Win10 on my lame little netbook which I only use for downloading Kindle books.
I don’t know how I managed to break Win10 – it won’t boot & cannot repair itself. My netbook has a Win7 re-install partition but I can’t seem to boot into it. Perhaps a lilo issue of some kind. It runs much better on my Slackware64-fvwm env anyway.
Rather than add multilib to my Slack64 env, using Slack live on a 4G memstick gives portable convenience which will work until Amazon forces use of a newer Kindle.
Again, thank you & best wishes. Another Bob 🙂
Hi Bob,
I have a Slackware Live USB stick on my keychain always, and it has been used on multiple occasions to fix computers. It is not a dedicated repair tool but that is not needed either.
I like the versatility of the Live OS and all its generic tools combined with a LUKS-encrypted homedirectory containing a copy of all my private stuff.
Eric,
I am interested in knowing if there are any companies that are using Slackware in an enterprise environment. I am completing a degree in Business IT, and since I really prefer Slackware, I would be interested in applying to such companies. Thanks so much for the great blog and your work on Slackware.
Benjamin, I can not answer your question. Companies using Slackware don’t come to tell me about it.
From time to time this question comes up on the LQ Slackware forum, my suggestion is that you do a search on the site and if you can’t find the answer, open a new thread.
Tried the latest live iso and I still run into the problem that I can’t log into my new Dell 13 7373 because the password just fills up with letters (or the bullets that stand in for passwords in KDM). So it boots to KDM, but can’t go anywhere from there. Same thing happened the last time I tried slackware current. Thought it was the kernel, but -current is back on a 4.14.X, iirc.
Joe, assuming that you did actually try the username/password combination of “live/live”, does the keyboard work at all? I mean, can you type the username “live” properly but is it the password entry field that is acting weird? will allow you to edit the boot commandline. Add a ” 3″ at the end and press ENTER to boot. If you have an UEFI computer you will have booted into GRUB. In that case, press the “e” to be able to edit the boot commandline, add ” 3″ at the end and press to boot the computer.
In that case, try booting into runlevel 3 (console) instead of the graphical runlevel 4. Specifying a different runlevel than the default of “4” should be done in the boot menu:
If you booted into the syslinux menu then pressing
I am curious to know if you can login properly on the computer’s console.
Then type “startx” to start an X session. If you want another than the default desktop, use the “xwmconfig” command at the console to select another desktop environment.
In KDM the user name is present. When I type the first letter in the password field it just starts filling in characters and the little circles continue to flicker as if the key was still being pressed. I tried both the log-in from console from KDM and changing the run level in grub and the keyboard didn’t work correctly to enter the user name. It only typed on letter for the l, but it was, iirc, a 3. The backspace got me a tilde (or apostrophe).
Joe, sounds like a keyboard mapping issue.
Exactly what ISO did you download, and how did you transfer this ISO to a medium?
Hi Eric,
i’m a long time user of Slackware and I thank you so much for all extra packages and scripts that you provide. My last dep-solution implies libinput, libxkbcommon e libwacom to use qt5 and qt5-webkit… to run TeamViewer. Big UP AlienBOB.
I downloaded the latest Slackware current live iso. No Mate or anything. I transfered it the usual way, the dd command. It works fine on my desktop, which has a 4th generation i5 and associated Intel chipset. It doesn’t work on the new Dell 13 7373, which has a Kaby Lake and whatever chipset comes with that.
Hi Joe & Bob,
Joe, if it were me, I’d try an external USB keyboard connected to the Dell before booting with the Slack Live iso. If things still hiccup then ???
Just a thought.
Regards,
Frederick
Downloaded the Plasma5 live iso, burned it via dd and it works fine. I’m on it right now. Not sure if that’s odd or not, but I’m definitely installing from the live disk while I’ve got it working. Thanks.
Joe, the Plasma5 Live ISO contains a more uptodate slackware-current, in particular a more recent kernel. Perhaps it’s the newer kernel modules that make your keyboard work with the Plasma5 Live.
Hi Eric, on a lighter side note, your Slackware Live distro is candidate at the LQ Member’s Choice Awards in the Live Distribution of the Year category:
https://www.linuxquestions.org/questions/2017-linuxquestions-org-members-choice-awards-126/live-distribution-of-the-year-4175620802/
Interesting 🙂
Thanks Frederick & alienbob. I installed -current today using a USB keyboard. The laptop keyboard works fine after the install.
Mr Alien,
What would you recommend to install on home-office laptop? Slackware-14.2 stable or slackware-current? I found your Plasma5 Live ISO running flawlessly on target hardware – good, old 2010 thinkpads. But I am not sure if stable branch can be upgraded to Plasma5 version.
I am Slackware server user since years. And now I had decide to get rid of “most-used operating system” on personal equipment.
Hi zycho,
If you are not an experienced Slackware user, I would not recommend to install Slackware-current.
You would be better off installing Slackware 14.2 and get a trouble-free security update process. The updates to -current are too frequent and rely on the enduser’s understanding of the OS.
You will not get the latest Plasma 5 with Slackware 14.2 because I stopped updating my Plasma 5 packages for Slackware 14.2 in November 2017 (only getting security fixes now) but even that November release is stable and should run smoothly.
In due time, Slackware 15.0 will be released and when it does, it will have Plasma5.
That was the formal answer, to keep you safe. If you want, you can of course install -current but it may require a bit more tweaking and at least it requires that you keep your -current installation uptodate with the updates that are released (sometimes multiple updates in a week).
Eric,
I’m sorry but I’m suffering from a terrible case of the dumb-****. I have slackware64-live-mate-current.iso and a partition on my hard drive that I’d like to put it on. I’ve created /mnt/livemedia, mounted the iso to it, and ran setup2hd and the “1. Mount your Live media partition on /mnt/livemedia.” error showed up. I then tried mounting my slackwarelive USB stick to /mnt/livemedia and got the same error. I’m running out of hair so can you point me in the right direction. The eventual partition on my hard drive is /dev/sdb2 if I can pull this off.
Thanks for all the work you put into slackware and your goodies. You’re amazing.
Jim
jim you actually have to be running the Live OS in order to use ‘setup2hd’. It looks like you only mounted the ISO but are running a regular OS off your hard drive. That is not going to work.
I’ll try that and report back. Sounds like a slight documentation thing maybe.
I did find a ‘feature’ of the “-r” option for iso2usb.sh. Seems that /etc/shadow is overwritten. It shouldn’t be a biggie to fix on my end but… 🙂
Thanks,
Jim
Well, the documentation states:
setup2hd
The fourth script:
The “setup2hd” script enables you to install the running Live OS to the computer’s local hard disk.
The issue with the “-r” option and overwriting /etc/shadow is weird. I have refreshed my own Slackware Live USB stick many times and the passwd/shadow files remain untouched.
Mine had a different ‘root’ password, no ‘live’ user, and a ‘jim’ user. The refreshed /etc/shadow has ‘root’ and ‘live’ with the default passwords. Interesting. I grabbed the iso2usb.sh from the link in the latest blog.
Would I be correct in worrying that my rc.inet*.conf and other network setups would be replaced with setup2hd? If so, is there an easy way to not copy the /etc stuff if it already exists on the target partition?
Jim
In Slackware 14.2 we can swap between kde4 and xfce by simply logging out, ticking a box and logging back in again.
In subsequent releases of Slackware, assuming it sticks with kde4, could there be an option to be able to swap between kde4 and kde5 by simply logging out, ticking a box and logging back in again?
Alex, the next release of Slackware will say good-bye to KDE4 and instead you will get KDE5. Or rather Plasma5 as it is called.
And regardless of how many Desktop Environments you have installed (even those packaged in 3rd party repositories, like LXQT, Mate, Cinnamon, Gnome3) you will always be able to select the one you want to use by using the session dropdown menu in the graphical login manager, or by using the ‘xwmconfig’ command if you boot into the console runlevel 3.
I installed chromium-64.0.3282.140-x86_64-1alien which seems to be working well but Help->About reports “Version 55.0.2883.87 (64-bit)”
Chris, in that case something’s screwed up on your computer. My Chromium – package from my repository – reports “Version 64.0.3282.140 (Developer Build) (64-bit)”.
OK, I’ll look around for odd files perhaps left over from previous (SBo) installations.
Alienbob, what’s going on? Nearly two years ago I asked if the Lumina desktop environment was going to make its way into Slackware and you replied it would not. Now I see you are maintaining a slackbuild for it. What happened?
As for the second question I asked two years ago, will Slackware ever switch to libressl? Ryan
ryan, the Lumina Desktop has not made its way into Slackware.
I don’t know where you looked, but it is merely a part of my own repository, that is all.
Just to check back in on my non-working keyboard issue with current on an 8th gen Intel-based laptop, I installed the latest iso and everything worked fine. Immediately upgraded to your latest plasma packages and it’s a very nice fit on this laptop.
Hi Eric,
I have upgraded your multilib package from
14.2 to current. The Mathematica program now
fails with the error
ZLIB_1.2.9′ not found (required by /usr/lib64/libpng16.so.16)
Is there anything I can do?
Will this always happen in the future?
Best wishes
Charles Nash
Charles – if you are actually running Slackware -current and have upgraded your computer with my multilib packages targeting slackware-current, then this error should not occur.
What did you download from what URL and what command(s) did you use to upgrade?
I uograded a mulit-lib enabed (as per your instructions) slackware 14.2 to slackware current using a mirror for the slackware-current iso (I didn’t keep a note of what the URL of the mirror was) and
I used the iso image do a full install. I then downloaded and installed the multilib packages using current as the version number, rather than 14.2.
Perhaps libpng is faulty?
Charles my assumption is that you did not correctly upgrade your computer from 14.2 to -current. You used an ISO image? There are 3rd party slackware-current ISO’s you can download (I host one of those myself) but how did you do the Slackware upgrade using that ISO? I am pretty certain that you missed one or more packages.
HI Eric,
I hope |I am not wasting your time. It is very good of you to get back to me.
To upgrade I simply booted on a usb stick onto which I had transferred the iso image using dd, allowed the installer to reformat the disk and did a full install.
Charles, use the command “ldd /usr/lib64/libpng16.so.16 |grep found” to find out if your computer is missing any libraries, and therefore packages.
Where did your download that ISO? If you don’t know the mirror you should still be able to tell me how you searched for and found that ISO.
Eric
that command, run as root, exited silently.
The library libpng16.so.16 does actually seem to be there.
To find the ISO I am afraid I took the lazy option of typing
slackware current iso
into google.
I conjecture that the latest version of this library has broken
something. Can I retreat to an earlier version perhaps. I say this
because googling for my error, just now, has turned up a comment from which one could infer this.
Please don’t spend too much time on this, though your expertise is greatly valued, and your advice received with much gratitude.
Eric,
I find that removing the
libpng
package that is installed by current
and replacing it by the one that is installed by 14.2 has solved the problem.
Charles
Oh Oh I spoke too soon. The original libpng error reappears after a reboot??
Charles, your issue is local to your computer. You are the only one reporting this error which should make you think.
Eric,
Would that that were so—-If one googles
ZLIB_1.2.9′ not found (required by /usr/lib64/libpng16.so.16)
there are quite a few hits and they involve the same problem: a program exits with the above error and, at least in some of the cases, a solution is to regress to an earlier version of the libpng library.
In my case this worked too, until I rebooted. I speculate that this has something to do with ldconfig; which I am afraid I know little about.
However, Eric, I apologise for drawing your attention to the possibility that your multilib packages were in any way responsible. The simultaneity of the error’s occurrence and
my installation of your packages looks, now, like a coincidence.
But thank you very much for your time, promptness, politeness and simple good manners in corresponding
with me.
All the very best
Charles
Charles if you ask me, you have a mix of packages that are taken from 14.2 and -current. The error cannot be explained otherwise.
Try this command:
ls -1 /var/log/packages
And post the output on a pastebin site (NOT ON THIS PAGE!).
Eric,
I tidied up the libpng libraries
and all is now well.
Over and out
I’m seeing an issue with the Python 3 parts of PyQt5.
The following works on Python 2 but not 3:
from PyQt5.uic.Compiler import compile
See here for Python 3 output: https://pastebin.com/Lp7f9Nc1
This is on Slackware64-current with your ktown repository.
Ben, interesting. In my PyQt5.SlackBuild script I follow what’s being used in Slackware’s PyQt 4.x script, which is to:
rm -rf $PKG/$PYTHON3LIB/PyQt4/uic/port_v2/
rm -rf $PKG/$PYTHONLIB/PyQt4/uic/port_v3/
… but now it looks like that breaks some functionality.
I can not find a good reason to remove these directories, so I’ll discuss it in the team.
OK I fixed this in git (https://git.slackware.nl/ktown/commit/?id=338b3b02b7bb7e3f8bde1645c694c2b6c6d42b11) and I will recompile the PyQt5 package as soon as I start working on the next monthly release for ktown.
Thanks for pointing out the error Ben.
Thanks Eric, I just rebuilt PyQt5 and it’s fixed.
Congratulations Sir. Slackware live edition came second on 2017 LinuxQuestions.org Members Choice Awards as live distribution of the year 2017. Running against live distros that are out for decades now. It came second by only 3 votes. A young project and so sucessfull already. Thank you!
fabio, I am glad people like my liveslak scripts and find the resulting Live ISO useful. It’s use has grown way beyond my initial expectations.
Por favor, acrescente a opção de espelhar o ARM também no script mirror-slackware-current.sh. Pretendo manter um espelho do Slackwarearm para trabalhar com o raspbarry Pi.
Muito obrigado!
———
Please add the option to mirror ARM also in the mirror-slackware-current.sh script. I want to keep a Slackwarearm mirror to work with the raspbarry Pi.
Thank you!
Antonio, I added ‘arm’ as a new supported architecture to the “-a” parameter. Please check if this does what you need.
Is there any way of using the 5.9 qt5.SlackBuild on Slack 14.2 ??
On mine it fails at the start with
In function `runMoc(int, char**)’:undefined reference to
`QCommandLineOption::setFlags(
QFlags)
Thank you.
Hi Eric!
Thanks for for all your work and all stuff you provide!
I noticed a problem related to “live streaming” from youtube using MPV player.
The problem was cache reaches “0 seconds” after 5 or 6 seconds after that the streaming is interrupted and mpv starts buffering. Then after many seconds of pause the stream is resumed, works for some seconds, then it is again interrupted while it buffering, and so on…
I solved thanks to slacky.eu community guys, by rebuild ffmpeg package after applying a patch from Arch Linux ffmpeg build.
Here you can find the patch I refers:
https://git.archlinux.org/svntogit/packages.git/tree/trunk/fs56089.patch?h=packages/ffmpeg
I used to use your ffmpeg package, rebuilt on my machine to enable patented libs.
So I tried the patch above:
– I put it, “fs56089.patch”, in “patches” subdir of your “build” directory:
http://www.slackware.com/~alien/slackbuilds/ffmpeg/build/
– then I edited your ffmpeg.Slackbuild (the function “make_ffmpeg”):
——————————–
make_ffmpeg()
{
echo -e “**\n** ffmpeg …\n**”
[…]
# patch to prevent pause/buffering on youtube (or similar) live streaming
cat $SRCDIR/patches/fs56089.patch | patch -p1 –verbose \
2>&1 | tee $OUTPUT/patch-${PRGNAM}.log
[…]
————————————–
I tried to rebuild “ffmpeg 3.4.2” with your Slackbuild and the above patch. MPV works good now, also with those types of live streaming video. So the issue seems solved.
You can test MPV with the following (it is sky news live youtube channel):
———————–
mpv –no-config ‘https://www.youtube.com/watch?v=XOacA3RYrXk’
———————–
As regards MVP I’m usign mpv-0.27.2 from SBo
———————————————
mpv-0.27.2-x86_64-1_SBo
ffmpeg-3.4.2-x86_64-1alien
—————————
After ffmepg rebuilt with the patch, mpv rebuild doesn’t seem necessary: the issue just disappears.
I’ve read of some user solved by upgrade to mpv-0.28, but it requests newer version of ffmpeg (git version i suppose). I didn’t try, but it is also possible that newer ffmpeg version are “patched” out of the box.
Hope this report can help, I don’t know if you plan to apply it to ffmpeg or just to try an upgrade of your package building a development version.
Thanks again for your work!
Bye! 🙂
PS.
The above live streamings were regularly played within the browser (I’m using chromium).
Hello Eric, thank you for the creation of additional slackware packages. One issue: Libreoffice 6.0.2.1.0+ 64bit consistently crashes immediately when I open “Tools” > “Options”. I cannot say, whether this is newly introduced in this version, as I have started to use Slackware just now. Kind Regards A. Schultz
A. Schultz, 32bit Libre Office? This issue was discussed in the comments section of a LibreOffice related blog post here.
Hello Eric, no, I wrote 64bit. It is Libreoffice 6.0.2 on Slackware 14.2 on a Thinkpad x220. I cannot untick Hardware Acceleration as Tools > Options immediately crashes Libreoffice without exception. Thanks in advance for any hints.
Sorry, correct, I have found the session, where it was discussed. It really is the secring.gpg file in .gnupg. Probably because og the new encryption feature in libreoffice.
Hi Eric!
Thanks for all your efforts on updating your slackbuilds for current after the 2018/04/19 “third ever Slackware Mass Rebuild”. Just a thought, is this a good time to add the “# Don’t ship .la files:” “rm…” command to your current builds? Your VLC3 04/22 build left /usr/lib64/libvlc.la /usr/lib64/libvlccore.la in the LD_LIBRARY_PATH.
Thanks again for all you do!
After reading something about DVDs not playing properly, i blew the dust off a DVD movie and tried to play it. No luck, regardless of what software package I tried.
With VLC the second I clicked “play disc” it crashed. Others, such as MPlayer, would access the dvd for a few seconds and then stop as if no command was ever issued.
So, just for chuckles I tried all the same players as “root.” and they all worked perfectly.
I’m guessing it must be a permissions problem, but I don’t know where to start.
What would you recommend?
Many thanks!
It must have to do with device files. Probably /dev/dvdrom or whatever the device holds the DVD has group access rights that your normal user is not a member of?
You could try strace to determine what devices/files the player is trying to open.
chuck56 – just remove those .la files from your computer please. I am not getting paid to deliver SlackBuild scripts or packages, so it will happen in my time. Whenever a new package needs to be built.
As I mentioned, VLC starts normally, but when I click on “open disc” it crashes with the following error:
ASSERT failure in QList::operator[]: “index out of range”, file /usr/include/qt5/QtCore/qlist.h, line 549
Aborted
Already provided recent feedback in LQ.org on the lack of [your? Pat’s?] good, clear documentation and support for GRUB as opposed to the default LILO bootloader in Slackware 14.2 and Slackware -current.
Am wondering why your VLC and VLC2 Slackbuilds ( http://www.slackware.com/~alien/slackbuilds/vlc/ and http://www.slackware.com/~alien/slackbuilds/vlc2/ respectively ) don’t directly list the same external dependencies as do Christoph Willing’s SBo VLC package at https://slackbuilds.org/repository/14.2/multimedia/vlc/ ??
While of course I’ve see the VLC Slackware site https://wiki.videolan.org/Slackware for complete VLC information here, that site isn’t referred to that well (or at all?) within your above pair of …/~alien/slackbuilds/ VLC links.
nycace36 – the vlc SlackBuild on SBo lists dozens of dependencies because it *has* dozens of dependencies.
My vlc packages (vlc and vlc2) *also* list all the dependencies that they actually have. The fact that my list of dependencies is so small is because I compile all these other libraries straight into the vlc package itself. They become internal libraries instead of external dependencies.
I am a bit surprised that you needed to ask. I have written a lot about how my vlc package statically links many libraries so that it is self-contained and runs with full functionality out of the box.
And why should I write Grub documentation for you? I already suggested that you can write this documentation yourself as a contribution to the community Wiki at docs.slackware.com if you find that it is important.
Mind you: Slackware is not my distro. I support the distro with time and effort for which I do not get paid. You don’t get to challenge me for a perceived lack of documentation support.
cwizardone I installed slackware64-current on a desktop with a DVD drive, installed VLC and inserted a couple of DVD disks. KDE popped up a message asking me what I wanted to do with the inserted disk and “Play with VLC” was one of the options.
They all played flawlessly – audio and video – with vlc running as my own user, and showed a proper DVD menu too.
Man, those DVD subtitles are ugly!
Alien,
Thanks for taking the time to check that out.
I use Xfce the vast majority of the time, where, as of just a few moments ago, VLC still crashes when I click on “Open Disc…”.
However, in -current with kde4 it does play as you described above and also plays perfectly when I click on “Open Disc…”.
Oh, well…. I can move over to kde4 when I want to watch a DVD.
🙂
Thanks, again.
cwizardone – I booted that desktop computer again, but this time into XFCE. I started vlc and through the menu, opened the DVD disk that I had inserted. The console output shows me that libdvdread detects the name of the DVD and then searches the CSS key it needs to decrypt the video. That succeeds and the video plays. Just like in KDE4. No crash.
Interesting.
I just switched back to Xfce and tried again with VLC. No luck.
However, the DVD did play in SMPlayer (MPlayer) in Xfce.
Who knows?…..
OK, this is interesting. I paused here and downloaded your
“U.S.” version of vlc-3 and it too crashed when I clicked on
“open disc..”
So I uninstalled it and downloaded and installed vlc2 from the “restricted” repository and it works perfectly.
Oh, well…… that will work when I want to watch a DVD.
🙂
Hi, I noticed your current multilib still includes libXfont and libtermcap even though current does not include those anymore. Maybe they should be removed?
orbea yeah I should remove those someday.
Good Morning (Your time). 🙂
I was just installing the “refreshed” multilib packages for -current, thank you very much, and noticed there were no “refreshed” packages for intermcap and libxfont and I don’t see them in the package list. Should they be removed?
Thanks, again.
cwizardone see the posts immediately above yours.
Hi alienbob, I am on Slackware current with multilib and KDE4 and have your vlc package installed. I am getting some problems when I play a video on slow motion. Specifically, when I jump ahead in the video, it freezes for a long time (10s +) until it (hopefully) continues. I ran from terminal and I got some error messages related to ffmpeg. In comparison, when I do this at normal speed, it works fine. I also have your ffmpeg package installed. The exact log from running vlc is:
vlc openaigym_video_train_episode_002500.mp4
VLC media player 3.0.2 Vetinari (revision 3.0.2-2-g34d098f437)
[00000000021cd310] main libvlc: Running vlc with the default interface. Use ‘cvlc’ to use vlc without interface.
TagLib: MP4: No audio tracks
QStandardPaths: XDG_RUNTIME_DIR not set, defaulting to ‘/tmp/runtime-niko’
qt.svg: /home/niko/.kde/share/icons/evolvere-black-folders-vivid/64×64/apps/vlc.svg:250: Could not resolve property: linearGradient5601
qt.svg: /home/niko/.kde/share/icons/evolvere-black-folders-vivid/64×64/apps/vlc.svg:250: Could not resolve property: linearGradient5601
qt.svg: /home/niko/.kde/share/icons/evolvere-black-folders-vivid/64×64/apps/vlc.svg:250: Could not resolve property: linearGradient5601
TagLib: MP4: No audio tracks
libva info: VA-API version 1.1.0
libva info: va_getDriverName() returns 0
libva info: Trying to open /usr/lib64/dri/i965_drv_video.so
libva info: Found init function __vaDriverInit_1_1
libva info: va_openDriver() returns 0
[00007ff70cc0b8c0] avcodec decoder: Using Intel i965 driver for Intel(R) Skylake – 2.1.0 for hardware decoding
[00007ff70cc0b8c0] main decoder error: Timestamp conversion failed (delay 576000, buffering 192000, bound 9000000)
[00007ff70cc0b8c0] main decoder error: Could not convert timestamp 178423701740 for FFmpeg
[00007ff70cc0b8c0] main decoder error: Timestamp conversion failed (delay 576000, buffering 192000, bound 9000000)
[00007ff70cc0b8c0] main decoder error: Could not convert timestamp 178430353161 for FFmpeg
[00007ff70cc0b8c0] main decoder error: Timestamp conversion failed (delay 576000, buffering 192000, bound 9000000)
….
Do you have any suggestions how I can fix this or debug it further. Thanks!
niko, perhaps the root cause is in the video file. Have you tried using ffmpeg to clean up the ml4 container and see if that helps (something like “ffmprg -i inputfile.mp4 -c copy outputfile.mp4” is the most basic attempt at that).
Else try playing with VLC’s cache buffer size. Or specify a different video-output. Or try disabling hardware decoding (libva log output shows that VLC activated its VAAPI support).
Hi Eric,
Love the work you do.
I apologize in advance if this is a stupid question but here goes.
Is it possible to get a program to run on Slackware 64 14.2 that has a dependency on libsystemd? I used to use synergy on my windows and slack boxes all the time but synergy2 has builds only for Debian, Ubuntu, and fedora/centos/redhat. are we just out of luck now for anything built for systemd?
– dr –
donrich39 – synergy2 is a closed-source program so you have run into the exact reason why us Open Source advocates promote the use of free software with the sources available. If this software has a hard dependency on systemd then Slackware itself is not going to be a supported OS, out of the box.
You can however look at Dlackware, which is Slackware fork with systemd and pam bolted on top.
I built a Live ISO for that team a while back, but I can not find if they still make it available online. They even removed their package repository. Perhaps you should send them an email.
Here we go again. 🙂
In -current:
******************
bash-4.4$ vlc
VLC media player 3.0.2 Vetinari (revision 3.0.2-2-g34d098f437)
vlc: Relink `/usr/lib64/pulseaudio/libpulsecommon-11.1.so’ with `/lib64/librt.so.1′ for IFUNC sock_getres’
vlc: Relink `/usr/lib64/pulseaudio/libpulsecommon-11.1.so’ with `/lib64/librt.so.1′ for IFUNC sock_gettime’
Segmentation fault
******************
Scratch the above. I still had vlc2 installed, so I uninstalled both and then installed vlc-3.02 and all is well.
Thanks for all your hard work.
🙂
Hi
wxBASIC package is missing in Slackware repo…
http://www.wxbasic.net/downloads
Can you please add it to official SW repo?
thanks
Also I found a bug in FRESH installation, which let us install Slackware without ANY SELECTED PACKAGE!
So this leads Installer stuck and getting bugged very bad
Here is how:
Put CD/DVD Slackware boot it to fresh install -> Follow the normal steps -> when Installer asking you to chose packages, UNSELECT ALL packages(even core) -> then press NEXT.
^ you will see that installer hangs and getting bugged
(sorry for bad english)
Retro – welcome to Slackware!
I think you still need to learn how Slackware differs from other distributions.
One:
If you want to use wxBasic and it is not found in Slackware and not on slackbuilds.org, then you can write a SlackBuild and submit that to slackbuilds.org. Don’t think that it will ever be included in Slackware itself. I am not interested in adding it to my own repository either.
Two:
Slackware assumes you know what you are doing and will readily accept whatever you tell it to do. If you select zero packages during installation then that is what you get. You are in control, and you are allowed to fuck up. Slackware is not going to prevent you from doing that.
Thanks for rebuilding the openconnect package on current. It appears podofo needs to receive the same treatment for the same reason.
Thanks again for your amasing work
I think the NetworkManager-openconnect plugin needs rebuild.
I seems to crash as its linked to older version of icu4c. Current has updated it and it fails to link with the new version.
Timo, I will recompile it and also the NetworkManager-openvpn package.
yesterday (june 28) I upgrade all packages with slackpack+ on my “current” slackware and multilib stop functioning.
attilio. Unless you make your statement more specific (i.e. add the actual error messages you encounter. and you tell me what it is exactly that used to work and now no longer works) I can not answer you.
Good day, I wanted to download slackware64-current-live from the mirror http://bear.alienbase.nl/mirrors/slackware/slackware-live/slackware64-current-live/ but found an error when assembling the image:
Write failed because No space left on device
FATAL ERROR: Failed to write to output filesystem
*** ./make_slackware_live.sh FAILED at line 1156 ***
— Cleaning up the staging area …
[20180718_0005 UTC] Finshed.
I’m sorry for my English
Hi Revorg. You seem to be starting the “make_slackware_live.sh” command which will download a full copy of the Slackware package tree and will then create an ISO image. That will consume many GB of diskspace in /tmp . Perhaps your /tmp is on a separate partition which filled to 100% ?
In any case, if you download an ISO image from the URL you mention, there is no reason to re-create that ISO from scratch using that “make_slackware_live.sh” script.
Hello Erik! I downloaded the iso you made for the latest KDE Plasma. It took hours to get the modules installed via setup2hd. Once finished I did a slackpkg update followed by a slackpkg upgrade-all and as I saw a new package for slackpkg+ I first upgraded that one before upgrading all the other packages. When I rebooted I had SDDM instead of KDM and I went into an infinite loop of logins. I had to reboot into recovery mode and I typed plasmashell. I had the following error message libphonon4qt5experimental.so.x (I don’t remember the digit x) no such file or directory. I had to choose another DE/WM with xwmconfig (Xfce). I had to put setxkbmap be into /usr/share/sddm/scripts/Xsetup to be able to log in with the correct password while choosing the Xfce session when I got the SDDM login screen back. When I was in the Xfce session I could download the phonon-4.9.1 from pkgs.org from your Ktown which replaced an older version you had in the packages of your iso (the 4.8.3). Then I could reboot and get the kdm login manager back. Could you remove that old phonon package so that people will be able to reboot correctly after doing the updates?
pourpier – something’s not right in your story. The Plasma5 version of my Live ISO contains phonon-4.10.0. The Slackware used in that ISO is Slackware-current. Slackware itself contains phonon-4.8.3. My Plasma5 ISO upgrades that to 4.10.0 which means the original phonon package is contained in one of the modules but is being masked out so that only the new is visible. When installing to harddisk using ‘setup2hd’, this process does not extract the individual squashfs modules, but it performs a rsync of the readonly Live filesystem. The old phonon package is not found there.
Also, KDM is not part of the Plasma5 ISO since KDM is part of KDE4.
It looks to me that you may have downloaded the standard Live Edition, containing a full Slackware-current, so with KDE4 and not Plasma5. This standard edition also comes with KDM whereas the Plasma5 edition comes with SDDM.
Apparently you started with a standard Slackware-current and somehow managed to do an incomplete migration from KDE4 to Plasma5. How you did that, I can not tell because the standard edition (the one with KDE4 and phonon-4.8.3) does not include slackpkg+.
And yes, installing through ‘setup2hd’ takes more time than a regular install from a Slackware installation medium. That is due to the additional layers of complexity in the Live OS (squashfs, overlayfs) that require CPU power and deteriorate the extraction speed of the packages from the Live medium to your harddisk.
Thanks for promptly updating the newer gcc packages for multilib, but can you also give this prompt treatment for llvm too? I’ve seen pcsx2 break in non-obvious ways in the past due to mismatched llvm versions and it might be nice to avoid this issue entirely in case someone else runs into it too. Thanks! 🙂
Hi orbea
I am in the process of building a new Plasma5 release, and with that comes a new PLASMA5 Live ISO, and for that I need an uptodate multilib package set anyway.
So yes: an update to llvm-compat32 (and others) is underway.
Thanks! Its much appreciated. 🙂
Hello Orbea. I have an different issue with pcsx2 like sdl2 cannot see my joypad. Root can see but as user no. I found it something wrong with /dev/input/eventxx. But do not know exatly how to solve this. Maybe you know the answer?
Thanks a lot for the updated qBittorrent package! Noticed that for some reason the qbittorrent.desktop file was missing – maybe just by oversight? No big deal, easy to extract from the package in the 14.2 repo.
Hello Erik
Appears there might be a problem with your libreoffice.SlackBuild script in the following section.
# Add KDE related arguments:
if [ “${ADD_KDE4}” = “YES” ]; then
KDEOPTS=”–enable-kde4 –disable-gtk3-kde5″
elif [ “${ADD_KDE5}” = “YES” ]; then
KDEOPTS=”–disable-kde4 –enable-kde5 –ensable-gtk3-kde5″
else
KDEOPTS=”–disable-kde4 –disable-kde5 –dissable-gtk3-kde5″
fi
What is –ensable-gtk3-kde5 supposed to be?
A typo Alex 🙂
Since Slackware does not yet have KDE5, this option is currently useless, but if KDE5 is going to be introduced, the “–ensable-gtk3-kde5” and “–dissable-gtk3-kde5” should of course become “–enable-gtk3-kde5” and “–disable-gtk3-kde5”
Hello Erik
Thanks for the reply. I had an idle moment this morning so I thought I’d take your libreoffice.SlackBuild for a test run
Isn’t your Ktown KDE5?
I’m keeping a VirtualBox instance alive with your Ktown updates on top of current and your multilib having ditched kde4 as per your instructions.
So in my case I thought the issue with libreoffice was not that “Slackware does not yet have KDE5” but that “libreoffice does not yet have KDE5” Or, if it does libreoffice is extremely fragile.
Correct me if I’m wrong.
Alex
Alex, my ktown is KDE5 indeed. And LibreOffice 6.1 has proper support for Qt5 and Plasma5.
If you remove those two ‘s’ the build should continue, although I have not tried this particular variation of libraries. NO idea if libreoffice compiles successfully against KDE5.
Hi Eric! I’ve noticed that the names of the recent qt5 packages for slackware-current use suffix “_alien”, instead of “alien”:
qt5-5.11.1-x86_64-1_alien.txz
Is this a typo or is it intended? I’m asking, because the new suffix doesn’t work with “[0-9]+alien” regex, which I’ve been using so far in slackpkg blacklist.
Also, I think this is my first comment here, and this is feedback page, after all, so… many thanks for all your hard work on Slackware! 🙂
tomac, I had to fix the qt5 package in my repository, its content was damaged (filenames with exactly 100 characters had lost the last character due to a tar bug). I did not want to release a “2alien” package because that would confuse users that have the qt5 package from the ‘ktown’ repository installed. Hence a “1_alien” build number.
That it messes with the blacklist line is unfortunate but it is what it is.
Erik,
A few years ago I got a git clone of LO no on 14.2, kde4, multilib to compile libreoffice. I keep it up to date to chase and maybe fix bugs – it continues to compile successfully.
It’s only recently that I cloned to current, multilib and kde5, qt5 and it’s also compiled successfully though the rendering was appalling making LO unusable. Lo say it’s work in progress. Having said that I compiled it again yesterday and there was a big improvement.
I’ve not yet managed to get a successful build with your SlackBuild script. I shall try again as I may have had cpu memory problems.
Eric, are you going to build a poppler-0.68 for the August Plasma5 release or wait and add it in the September release?
Hi Eric and thanks for the vlc-3.0.4 upgrade, works flawlessly as ever. Just a little surprised to see that the upgrade created two .la files in /usr/lib64: libvlc.la and libvlccore.la
I was under the impression that .la files in this library were seen as obsolete (Pat’s wording in the Big April Upgrade changelog post). But maybe I’ve misunderstood something?
Yeah I had forgotten to check that. I have updated the vlc.SlackBuild and next package update will not have the .la files anymore.
Thanks!
hi. many time ago i suggested the libfdk_aac encoder on the ffmpeg package. Thank you for tooking your time and indulging that. Recently i acquired a nvidia card and of course went to test gpu encoding/decoding. Your package worked like a charm. Would it be possible to compile it with support for libnpp to make it possible to use a gpu based scaler? Thank you for your time and patience.
ps: please persuade Volkerding to use your ffmpeg on slackware 15 🙂
fabio, adding support for libnpp would require that I add the Nvidia binary drivers as well. I can not enable support for NPP in ffmpeg otherwise.
That is not a desired solution, because it will force everybody, not just Nvidia GPU owners, to install these Nvidia binaries.
I am afraid you’ll have to recompile the ffmpeg package yourself.
Thanks for the very quick reply, i’ll do so.
On another topic, i use your veracrypt package for slackware and there is a new version out. Would it be possible to have it packaged for slackware? Again thank you for your efforts.
Eric,
Almost a year ago I did a clean install of 64-current. I had been running 64-current since its inception. I did not at that time go back to KDE5. A few days ago, I did a clean install in a VM (Virtualbox) and then replaced KDE with your packages.
Using Slackpkgplus, I first removed KDE.
Then I uncommented the Pkgs Priority and repo for KDE5
Ran Slackpkg update
Ran Slackpkg install alien-kde
Ran Slackpkg upgrade-all (for some reason, their were upgrades to about a dozen kde pkgs)
Ran xwmconfig
Reboot
Done.
This was a bit simpler than using rsync from the README of which it didn’t work anyway.
Then I ran it again on my regular machine and it went flawlessly. Total time for either was about 42 minutes.
I guess I was just wondering if this method is kosher, or did I just get lucky…..
Oh yea – this desktop is awesome! Thanks!
Hello, alienbob I’m testing liveslak 1,3 and I have some serious problems with the sound out, I will test for hardware problems first
Michael, the process of upgrading to KDE5 (Plasma 5) using slackpkg with slackpkg+ *is* simple. I use it all the time.
The “slackpkg install alien-kde” is the crucial step.
Hello : thanks a lot for your big packages, that I am too lazy to compile !:-).
I still have problems with the sound, it just will play for hours and then stop, nothing is coming out from the speakers
Hi Eric,
I’ve been using liveslak for a couple years now (thanks for helping me back with getting wine running!) and I’ve got another issue I could use help with:
One of my main use cases for liveslak was a second boot option for my corporate laptop. Worked great as a way to use linux and not touch the hard drive.
My new hp laptop has secure boot enabled and the bios is locked so I can’t disable it – liveslak won’t authenticate. Is there a way to get liveslak signed to work on secure boot machines?
Thanks much,
JeffB
JeffB – liveslak does not support SecureBoot. No way around it except by paying a Microsoft fee.
Hello Alien I figure out the problem with the sound, i was bad cable connection, but sometimes when I install slackware 15 I’m keep gettin a screen with 999999999999999999999999999, I wonder what happens
There’s no Slackware 15.
The “L99 99 ..” error is caused by an incorrect configuration & installation of lilo. Search the internet or post on a Slackware forum to find answers.
Hi, Bob.
I recently ported slackware64 ( based on current brach ) to aarch64, armv7hl and even mipsel ( mips creator ). As I’ve create seamonkey package for aarch64, I think it’s time to release it.
My job needed to have several SBC ( RK3399, alwinner, friendlyarm, etc ) running linux with native LCD and touch support (no raspi), so a low entropy distribution was the perfect solution.
I´ve done this filtering the original SlackBuild scripts trough bash, rather than create new ones, as slackwarearm do. This allows, for example, to change a gcc switch in one file and rebuild the entire system, or add a new arch with a few lines of code.
The package include a cross root filesystem creator (slackroot) and a package creator.
This means you can run slackroot on amd64 to bootstrap a aarch64 slackware compatible root filesystem, and the jump to it on a SBC or a qemu, and finish the packages. This is based on your work.
The result scripts keeps in mind the possibility of cross compile, dependences (slackbuild kcalc , for example will lauch the build and install on both runtime and dev dependencies ) No install dependencies at this moment, because of absolutely compatibility with original packages, but can add easily.
linuxdoc-tools have been especially challenging, because they taited the build root filesystem, and this is not very covenient for cross compiling. A pain, for short.
I woul like to share this work, and finish it. What I ask from you is some advice to share it, and a good site to publish it. I think this port can give Slackware further years of popularity.
At this moment lacks some installers, (must be done by hand) because of the miriad of SBC but this has solution.
My backgroud is from electronics (I even have won an award from ELEKTOR when 16 ) and microcontrollers, and this is the reason I like Slackware. I even got a Job request from a guy called Florian Damian ( from Facebook ) because my wide background, very tempting (especially wage), but no ethic. LOL.
Neverless, tank you for your cool work on slackware.
I can send you some photos of some working SBC, controlling Allen Bradlley PLC (which protocol I’ve implemented). Pretty
Regards:
Caso. (JACS)
Hi, Bob.
I recently ported slackware64 ( based on current brach ) to aarch64, armv7hl and even mipsel ( mips creator ). As I\’ve create seamonkey package for aarch64, I think it\’s time to release it.
My job needed to have several SBC ( RK3399, alwinner, friendlyarm, etc ) running linux with native LCD and touch support (no raspi), so a low entropy distribution was the perfect solution.
I´ve done this filtering the original SlackBuild scripts trough bash, rather than create new ones, as slackwarearm do. This allows, for example, to change a gcc switch in one file and rebuild the entire system, or add a new arch with a few lines of code.
The package include a cross root filesystem creator (slackroot) and a package creator.
This means you can run slackroot on amd64 to bootstrap a aarch64 slackware compatible root filesystem, and the jump to it on a SBC or a qemu, and finish the packages. This is based on your work.
The result scripts keeps in mind the possibility of cross compile, dependences (slackbuild kcalc , for example will lauch the build and install on both runtime and dev dependencies ) No install dependencies at this moment, because of absolutely compatibility with original packages, but can add easily.
linuxdoc-tools have been especially challenging, because they taited the build root filesystem, and this is not very covenient for cross compiling. A pain, for short.
I woul like to share this work, and finish it. What I ask from you is some advice to share it, and a good site to publish it. I think this port can give Slackware further years of popularity.
At this moment lacks some installers, (must be done by hand) because of the miriad of SBC but this has solution.
My backgroud is from electronics (I even have won an award from ELEKTOR when 16 ) and microcontrollers, and this is the reason I like Slackware. I even got a Job request from a guy called Florian Damian ( from Facebook ) because my wide background, very tempting (especially wage), but no ethic. LOL.
Neverless, tank you for your cool work on slackware.
I can send you some photos of some working SBC, controlling Allen Bradlley PLC (which protocol I\’ve implemented). Pretty
Regards:
Caso. (JACS)
altomaltes – that’s a nice achievement! I did an ARM port long ago (https://alien.slackbook.org/blog/armport/) but never finished it with a proper installer or installation images, although it ran nicely on my Chromebook.
If you want to release your work, there are several possibilities.
First of all, the sources and scripts you used, did you keep track of their progress in a version control system like git? If so, then you could just import them all into a public git repository. Like github.com, or gitlab.com, or on git.slackware.nl.
If you did not do any version control and just have a set of scripts, documentation and other files, you can (and should) _still_ import those into a (git) repository. That way, other people can create enhancements and feed your their patches.
About storing images, packages, installer, kernels etc, I think that the slackware.uk site is the best option. Darren Austin who is the site maintainer, keeps a lot of projects and mirrors alive on that server, and it has great bandwidth.
Very important is that you properly document your project and how you achieved what you did. You can do that in text files that are part of the sources & scripts, but you could also write one or more articles in the Slackware Documentation Wiki (ARM section): https://docs.slackware.com/howtos:hardware:arm:start
Let me know if you need help with any or all of this. Use my email address (alien at slackware dot com) for that. Works better than blog comments.
Hi alienbob, excuse me if this is out of place. I’m running Slackware64 -current and some packages from “alienbob” repo (not restricted) keep appearing as upgrades to native Slackware packages, they are id3lib-3.8.3 (same version in Slackware), jansson-2.7 (Slackware version is 2.11) and ninja-1.7.2 (Slackware version is 1.8.2). There is another package, recordmydesktop-0.3.8.1 (same version in Slackware) but it is in ‘extra’ directory in Slackware.
Hi Paulo
I may have to remove some of those packages from my -current repository indeed. I hope to find some time for that during the weekend.
Thanks for reporting.
Hi Eric
would you consider upgrading the restricted ffmpeg package to version 4 for slackware-current? it seems to be needed by latest versions of some packages in SBo (mpv in my case), for the moment i’ll try to compile it myself, maybe there are other options too that i’m not aware of.
Hi Matteo
The fmpeg 4.x releases are still incompatible with a lot of other software, so I am not going to upgrade my 3.x package to 4.x soon.
Hi Eric,
thanks for the reply, in the end i compiled it from SBo including the non redistributable libs, the package is made to coexist with ffmpeg3.4 and everything is working very well so far. It didnt take much time either, sorry for the dumb request.
Looking forward the next kde update!
Hi Eric,
Can you please add graphite2 to your multilib repo on current? It will be helpful for PCSX2 users which now requires a multilib version on current.
PCSX2: error while loading shared libraries: libgraphite2.so.3: cannot open shared object
No problem orbea. Will do that.
Hi, Eric. I think your pages are very good here, but I’ve a few
observations.
Re ‘Privacy Preferences,’ it sounds good but if ‘I Agree,’ what
am I agreeing to? ??
I don’t see a 3-letter username here. It’s an old idea and in my
view, still a good idea. Everyone knows rms and esr, how do
you feel about e?h or …something?
I’m here now because I’m thinking about a usb plugin with
‘persistence’ that runs a live Linux, and how to do that??
And because I’m concerned about what’s going on with
Slackware? Is Eric Hameleers the new Top Person in
Slackware Linux? I think you sure are a strong candidate
for it. Re which, smiles and Thankyous are good but these
do not pay those bills etc. I’d like to find an uncomplicated
way to send you some money, doing this acknowledges the
reality of doing things. If I can simply mail you a personal
check, it’s what I’d like to do. ??
Thanks — Martha Adams
THANK YOU !!!!!not just all your work for slackware
and thanks to you I at last have a shinny new handbrake
package for slack ware
same rob.rice on linuxquestions
Hi Martha
Sorry for the delay in answering but I have been in bed for the past two weeks, due to a pelvic nerve condition which prevents me from sitting or walking. And using a computer is quite painful.
It’s the European GDPR law (https://en.wikipedia.org/wiki/General_Data_Protection_Regulation) which demands that every web site that is being served to European citizens tells its visitors what personal data is being collected and to what purposes and extent.
This blog’s ‘privacy policy’ page (https://alien.slackbook.org/blog/privacy-policy/) informs you about this. And the popup asking for your ‘OK’ is meant to make you aware of it.
Yet, as this is a simple WordPress blog and no one has an account here except me, all that the blog collects is names/nicks and email addresses – and those are not even required to post a comment (although I hate to respond to ‘anonymous’ posters).
I don’t really get what you mean by 3-letter user names. Forty years ago, a sufficiently small amount of people had access to computer networks that 3-letter usernames would cover the needs (mine used to be ‘eha’ at the time) but with billions of connected people, 3 letters are not going to cut it.
About creating a persistent live Linux USB stick, my landing page https://alien.slackbook.org/blog/slackware-live-edition/ should cover a great deal of what you need to know if you were to pick Slackware for that goal. I have a 32 GB USB stick with a persistent Slackware Live (Plasma5 variant) attached to my key chain and I refresh its live modules everytime I release a new ISO image (quite easy to do, read the docs).
I am not the new top person of Slackware, haha. Pat Volkerding is still the BDFL and he will not be resigning. I do my part too, mostly focusing on KDE Plasma, multilib and the bigger packages that are not easy to compile. Plus documentation and support of course.
About monetary support – Slackware is my hobby, not my profession. Only because it eats up a lot of my (family-) time working on Slackware in the evenings, am I happy to receive the occasional Paypal donation as a show of appreciation. The Paypal donations also pay for the domain names, the server hosting and the physical (build-) servers that I need for all of this.
So, I would rather that you send a mail check to Patrick’s home address instead of mine. As long as we all can keep Pat’s financial situation healthy (donations are currently his only source of income), we will be able to keep Slackware alive.
Hello Eric, small bug in the changelog generation.
In http://www.slackware.com/~alien/slackbuilds/ChangeLog.txt
+————————–+
/bin/bash: Date: command not found
openjdk: updated to 8u191_b12 for Slackware 13.37 and newer.
Georgi – that was a manual error, and I will fix it, thanks. The script itself is working OK.
Hi, thank for your work. I see that you update really fast your usb-iso. I tried to install it (your iso, put on a usb key with dd, etc.), but the setup file was not running. I don’t know if it’s me or not, but the setup file is working with zenwalk + unetbootinusb.
Lucky, what USB ISO are you referring to?
Hi Eric!
I browse with uBlock Origin installed.
However my adblocker is explicitly disabled for your site.
I noticed that you put at the footer of your page a notice saying: “Dear visitor, you seem to be using an Ad Blocker. Please consider whitelisting Alien Pastures. I use the revenue from displaying ads (small as it is) to keep this site running. Thanks!”
Be advised that this notice still displays when all adblocking is disabled for your site.
Thanks!
Hi Eduardo,
Thanks for reporting that, and also thanks of course for whitelisting my site,,, I will have to investigate why de-adblocker still shows the message when you whitelist the site in uBlock Origin.
Thanks Eric!
It’s a pleasure to whitelist your site. The ads are always tasteful and often relevant. I even clicked on them many times because they surely interested me. I’m really grateful for being able to support you in this very small way.
Today I don’t see your footer banner. Whatever you did seems it was right. Thanks!
Eduardo, I removed the de-adblocker plugin and installed ad-blocking-advisor. It seems to do a better job.
Indeed. Thanks again Eric!
I read on LQ that you were having some health issue.
I wish you a speedy recovery.
Health and safety first. Package maintenance can wait.
I enjoy and learn a lot reading and understanding all your scripts. I use sometimes “portions” of them in my own scripts. Is it enough to mention you in my scripts? I do not want to use code from you without your permission. Regards and I hope you feel well now regarding your health.
Hi Francisco,
Most of my scripts are “BSD Licensed”, you will find this license statement in the script’s header.
The BSD licence allows free re-use of a script’s code.
If your modification re-uses large parts of the original script, you are obliged to keep the un-altered license text in your version of the script.
If you use only small chunks of the script in another script which has a different flow and function, then you are free to use that, and don’t even have to mention where you got it from. Mentioning my name is only a friendly gesture, but it is not mandatory.
Good luck!
Hi Eric, just wanted to let you know that it seems like slackware.nl is down for the moment.
Thanks!
Eduardo, it looks fine from here. Perhaps it was only a temporary glitch in the network.
Indeed. All normal now. Thanks Eric!
Hi Eric, thanks a lot for the upgraded Wine package. I followed your advice in the changelog and installed OpenAL, and also converted the 32bit OpenAL to a -compat32 package.
Most important for me is that Wine allows me to run Adobe ADE 4 (which was a pain in the ass to install and get running). When starting ADE for the first time after upgrading Wine I got a bit nervous when, for a few seconds, an error message appeared saying that some rundll.exe file couldn’t be found. But the message disappeared, and then everything ran just as it should.
Hi Alien Bob,
I have a question regarding the checksum algo used with Slackware. Would it be feasible to move to something like Sha256 instead of MD5SUM?
Just asking since the latter is considered only semi secure.
regards,
Amith
Hi Amith
The checksums are not used to determine whether a package has been tampered with. It is just a means for quickly determining the integrity of your local copy of downloaded files.
We use GPG signatures to validate whether a package is unaltered since its creation.
As for the use of MD5 versus SHA256 – I follow the format of the Slackware repository metafiles. As long as that uses MD5, I will do the same, to ensure full compatibility with the Slackware repo tooling.
There was a thread on LQ on how the glibc-i18n -current upgrade of Feb 1 resulted in a very large /usr/lib/locale folder (>900MB, an increase of ~700MB). A new upgrade came Feb 17 with the message “Go back to using the glibc.locale.no-archive.diff patch rather than the new built-in no-archive locale target. Avoiding hardlinks is not worth 700MB of useless bloat. Thanks to baldzhang.”
Since I use multilib I of course don’t upgrade glibc* from Slackware but instead use your multilib versions. I upgraded to your glibc-i18n-2.29_multilib-x86_64-3alien.txz (Feb 17). Now I can see that my own /usr/lib/locale folder contains 5856 files in 977 folders with a total size of 921.4MB.
I’m not running out of storage, so no crisis, but it makes me wondering if something went wrong with my upgrade?
I have no idea, I applied the same changes to my SlackBuild as Pat did to his, then rebuilt the glibc packages and the version ‘3alien’ is the result.
OK, thanks… as already said, no big deal as I’m not running out of storage.
Tried a reinstall, but no changes.
Is it just me? Do you yourself (or anyone else) have the multilib version of glibc-i18n installed? If so, what’s the size of your /usr/lib/locale folder?
Hi Eric,
Would it be possible to add compat32 packages for xorgproto and libpthread-stubs? I found they were needed to build the 32-bit portion of wine-nine-standalone with a multilib setup where the pkg-config dependency checks failed for x11.pc otherwise.
https://github.com/iXit/wine-nine-standalone
Thanks!
orbea, I was going to look into wine-nine-standalone (in fact I have a SlackBuild for it already) so it is good to know that that has new dependencies.
I’ll add them.
Thanks, I also submitted a SlackBuild to SBo without realizing you were also looking into it. 🙂
I can only test with radeonsi, but it works here.
I think SBo should have it regardless of whether I will put up a package. I don’t think I can test it myself.
Hi Eric,
This may be a trivial issue, but i am trying to download your latest multilib, and i am hit with the claim that slackware.nl uses an invalid security certificate:
“The certificate is not trusted because it was signed using a signature algorithm that was disabled because that algorithm is not secure.”
Is that a real issue?
jonivar:
https://alien.slackbook.org/blog/adding-cacert-root-certificates-to-your-slackware/
Hi alienbob. I’ve been using slackware64-current for a while. Never found the correct way to send feedback.
Several apps in 64-current are linked to libsasl2.so.2 (such as eg. mplayer) but the sasl2 version on the collection in .so.3. A simple symlink solves the issue, but it doesn’t seem correct.
Hi John
Sounds like you have one or more outdated (perhaps older custom-compiled or 3rd-party) packages installed.
The mplayer in -current is actually (correctly) linking against libsasl2.so.3.
Hey alienbob today after I make my system I tried to compile 32bit package of soundtouch after I did 32dev.sh profiling stil iin and I got a x86_64 pkg. I wanted put this in linuxquestions.org but for some odd reason I can’t find button to add new thread.
Thnx in advance
Renars – can you show the *exact* commands as you typed them in order to compile that package.
About linuxquestions.org: if you do not login with an account you can not create new topics.
In linuxquestions after log in I can’t see but can make blog entries.
But here is commands #. /etc/profile.d/32dev.sh
ie get promt like { root@darkstar (32bit):/home/username/slackbuilds/app}
then i do sh ./*d and process is rolling out but in and I get /tmp/soundtouch-1.9.2-x86_64-1_SBo.tgz
Renars, you overlooked the part of the instructions that tell you to define the ARCH variable and set it to “i586” or another 32bit architecture.
If you do not do that, then ARCH will automatically be set to “x86_64” by the SlackBuild script.
alienbob Yeah I really missed that part thanks a lot
Hi Eric, saw in your Changelog that you updated icu4c-compat to 64-1 but it does not show in your mirrors (I can see only 63.1). Did you pull the update?
Thanks!
Eduardo, the new package is in the repository actually, but mis-placed in a subdirectory. The slackpkg+ would pick it up but looking yourself you would probably not find it.
I’ll fix the repository.
Indeed I was looking for it by hand. Thanks again Eric.
Is icu4c-compat necessary anymore? I saw in Slackware’s changlog that Pat added older versions to aaa_elflibs:
$ fgrep /libicu /var/log/packages/aaa_elflibs-15.0-x86_64-6
usr/lib64/libicudata.so.63.1
usr/lib64/libicui18n.so.63.1
usr/lib64/libicuio.so.63.1
usr/lib64/libicutest.so.63.1
usr/lib64/libicutu.so.63.1
usr/lib64/libicuuc.so.63.1
Of course, your package contains some older icu versions besides these.
Ricardo, the icu4c-compat is not required at this moment in time, but pay attention to the full sentence in the ChangeLog:
“Temporarily added: libicudata.so.63.1, libicui18n.so.63.1, libicuio.so.63.1, libicutest.so.63.1, libicutu.so.63.1, libicuuc.so.63.1.”.
Those libraries were mainly added to prevent Plasma5 users from getting a nasty surprise. They will be removed eventually from aaa_elflibs.
An issue with upslak.sh:
after I changed the init on working USB stick with liveslak on it with the standard template (upslak.sh -i liveinit.tpl -o /dev/sdb), the stick can’t complete the boot.
With ‘debug=2’ in boot parameters there’re things in diagnostics about /sbin/init not being available, tty inaccessible, etc. The shell prompt is then offered.
Without the debug parameter, the problems seemingly start after init outputting “Writing persistent changes to media directory /persistence”, and there’s some complaining about “lowerdir=” set incorrectly or empty.
From the shell prompt, “…/mnt/live” and “…/mnt/liveslakfs” look like they are mounted (correctly? I don’t know).
Yury if you replaced the init script with one that you wrote (or modified) then that is where you have to look for the root cause.
I can not help you without knowing what you did.
Actually I think I see a possible root cause, if you are using the standard template file unmodified then upslak.sh fails to replace some of the template strings that were recently added (SQ_EXT_AVAIL, DEF_KBD, DEF_LOCALE and DEF_TZ).
I will have to update the upslak.sh script to fix that error.
Yes, the issue occurs even after replacing with the standard liveinit.tpl (so the stick behaviour was supposed to go back to normal). Thank you!
Hi,
I personally follow the mesa git master and create my own mesa compat32 package after compiling it in a 32-bit chroot. Recently upstream has deprecated their autotools build and now has broken it for me while I am told it is soon to be deleted in favor of meson.
Given this I switched my build processes to meson which led to the discovery that mesa now links against lm_sensors. I don’t think this effects current yet, but maybe it would be good to either keep in mind that a compat32 lm_sensors will be needed in the future or to preemptively add it?
Hi orbea,
I guess it won’t hurt to add lm_sensors already,
Yury, I have updated the upslak.sh script with the changes made recently to liveinit.tpl and uploaded the ‘liveslak’ script updates (to the web and git servers). The updates should fix your issue.
Eric, I’ve tested the new upslak.sh, and it works okay (although requires re-making the ISO and re-writing it to the stick). Well done, thank you!
Hi,
I noticed you do not include a compat32 package for pkg-config and since wine-nine-standalone would be using i586-slackware-linux-gnu-pkg-config for 32-bit/multilib builds it might be needed. I accidentally missed this by using pkgconf instead.
However when I tested a compat32 pkg-config package using convertpkg-compat32 I found it did bad things to pkg-config’s doinst.sh and left only a single line.
mv usr/lib/pkgconfig/$movefile usr/share/pkgconfig 1> /dev/null 2> /dev/null
I’m not sure how exactly this worked as I did not notice until later, but I found my /usr/lib/pkgconfig directory was relocated to /usr/share/pkgconfig/pkgconfig.
I think this is at least worth pointing out.
As for wine-nine-standalone it might be better to just use pkgconf instead, but I’ll have to consider it more.
After some more testing I was able to build wine-nine-standalone with pkg-config.
I think the options are to either document that multilib users should use pkgconf which can solve this with the provided shell wrapper scripts or to provide a compat32 pkg-config by default.
To do this I removed the faulty doinst.sh and moved i586-slackware-linux-gnu-pkg-config to /usr/bin/ before manually recreating the package. I am not sure what steps would need to be done to convertpkg-compat32 to achieve this. Using /usr/bin/32 lead to meson to infinitely hang on /usr/bin/32/cc…
I am happy to go with what you suggest is best. 🙂
You are probably aware of this, but if not, just FYI, the link to the multilib change log in the panel on the right side of your blog page now says, “Unknown Feed.”
cwizardone hmm yeah, I had to move the multilib repository out of http://www.slackware.com because its harddisk was full… I’ll fix the RSS url, thanks.
Hi Eric. This morning I’ve upgraded my system with slackpkg like I do always. After finished the system become very slow. I’ve reboot and when it finish start SURPRISE !!! X doesn’t work. The error message said “lebEGL.so.1 not found”. The problem seems to be mesa 19.0.4 version. This file is not included in the package. I’ve downloaded mesa 19.0.3 X works fine. Any news about this change in mesa package?
Gerardo, did you do a “slackpkg install-new”? I think you did not install the new mesa dependency “libglvnd” which is where you would find that missing library now.
Happy belated birthday. Thanks for all your efforts re: Slackware. You have helped me keep Slackware running since I switched over exclusively many, many moons ago. I will not begrudge you the occasional rest for family and health. -Ed
Hey Eric, today i built everything for laptop-mode-tools and after installation and adding it to /etc/rc.d/rc.local it wouldnt work heres terminal output
bash-4.3# /usr/sbin/lmt-config-gui-pkexec
Exec with pkexec
python3: can’t open file ‘/usr/share/laptop-mode-tools/lmt.py’: [Errno 2] No such file or directory
And it’s because the file Lmt.py has not been chmod +x so thats all thats wrong with it thanks
Hey sorry its because it needs chmod’d and the script in /usr/sbin/lmt-config-gui says
python3 /usr/share/laptop-mode-tools/lmt.py
but the file in /usr/share/laptop-mode-tools/ is named LMT.py
hopefully you get this message, didnt want to make a thread at LQ
Linus72, I fail to see the relevance for this blog.
I do not maintain a laptop-mode-tools package anywhere.
Please *do* create a new thread at linuxquestions.org for this, or post to the slackbuilds-users mailing list to get feedback for your issue. I am not your personal troubleshooter.
Eric,
I’m having issues with multilib-9.1.0 when compiling. It’s mostly stupid things like ‘configure’ not finding libs when they’re in plain sight (/usr/lib64 and /lib64). Is there any chance that you still have the packages for multilib-8.3 and the compat32 packages. If so, could you find it in your heart to send me a link to them? I’m hoping that reinstalling them will make things better until I can dedicate a day to trying to figure out what’s really not working with 9.1.0 and go from there.
Thanks,
Jim
Hi Jim
This must be an issue on your end. I have compiled quite a lot of software with the multilib gcc-9.1.0 by now, and have not run into the issue you describe.
I do not have older versions of my software ready for download, so the next best thing you can do is to go back to the Slackware non-multilib gcc 9.1.0. That will not affect your ability to run 32bit programs, you will just not be able to compile 32bit programs easily.
I did revert to 8.3-non-multilib and can’t get wine to work. I’ll play with it more to see why and not take up bandwidth here. I’ve read a few reports of libraries not being found that others have submitted and when I tried to start Xwindows Mate on my laptop, it complained of a missing lib. Changing to Xfce, starting it, exiting, and changing back to Mate fixed that one. I wish I was more versed in the compiler environment setup so I could provide more detailed info. Until then, cheers and thanks for all you give to the Slackware world.
Jim
Hi Eric,
I just noticed your mkvtoolnix package no longer works on current. It seems a version bump to 34.0.0 would make it compile fine btw.
Thank you for all your work !
Regards
I have a question about the ‘unofficial’ Slackware git, in particular the sysvinit-functions package: https://git.slackware.nl/current/tree/source/a/sysvinit-functions
The repo doesn’t contain the source package, i.e. function-src-8.53.tar.bz2 , which would have been understandable if is was an ‘independent’ package. It is not, however. That is, although the slack-desc says:
“This script was taken from Fedora”, at the moment Fedora is already quite far from this code.
Or would you imply that we should download it form slackware-source directly?
I mean, the package is not big.
lockywolf, let me tell you a bit about the reason for having this git repository. It tracks the changes in the sources, scripts and documentation of *Slackware* specific content. This is only to make it easier for everyone to debug their issues when an update to slackware-current breaks their system.
If you look through the repository you will find *no* trace of any 3rd party source tarball. I do not track the remote sources in this repository. If you want to make a change to an actual Slackware package, please always refer to the official Slackware source directory tree. in casu, http://ftp.slackware.com/pub/slackware/slackware64-current/source/a/sysvinit-functions/ . The best place to suggest an update is in the LQ thread https://www.linuxquestions.org/questions/slackware-14/requests-for-current-14-2-15-0-a-4175620463/
As you said, this git repository is an unofficial one. It was created and is being updated by an automated script running on one of my servers. It uses the official Slackware source tree as its origin. You can not request any modification of this git repository’s content.
Hi Eric,
First, thank you for all your work.
I have a question about some packages. I noticed that some of them, like potrace could be in conflict with their versions on Slackbuilds.org (“alien” is 1.10, “SBo” is 1.15), which could lead to break other packages. For example if I set potrace as an external dependencies of Inkscape (SBo), it will break at the next upgrade. I experienced other breaks with ffmpeg (link to x264, but can be due to an explicit version set from me) and python3.
Of course, these breaks can be spotted quite easily and I’m supposed to know what I am doing when I compile and install a package. I just wanted to know if you published some recommendations for people using both SBo and your repositories (for example a blacklist of some SBo packages if we want to use your repos).
Regards,
Moun
Hi Moun.
I do not have any recommondation. You are supposed to figure it out yourself I guess. The scripts at slackbuilds.org are submitted and maintained by hundreds of people. My own packages are maintained only by me. It so happens that if I do not use a package myself, or if I am happy with a version of the package, I will not update it in my repository. That can cause conflict with SBo and other repositories, yes. But I am not going to spend time monitoring & tracking what happens on SBo and other repositories… my time is limited.
>>through the repository you will find *no* trace of any 3rd party source tarball.
This is NOT a 3rd party source tarball. Its relation to Fedora is distant at best.
Hi Eric,
“my time is limited.” I think I can understand very well this sentence. 🙂 I just wanted to know. Thanks again for your contribution.
Regards,
Moun
lockywolf I think you are totally missing the point. But I am not going to discuss it further. If you want a Slackware package update, do not mention it here on my blog.
Hey Eric,
Thanks for all your amazing work. You are one of the main reasons I am still able to stay on my favorite distro Slackware. I am trying to install your QT5. I am on current. But I get undefined reference to `std::__cxx11::basic_ostringstream<char, std::char_traits, std::allocator >::basic_ostringstream()@GLIBCXX_3.4.26′
as part of my errores while running the build. Do you have any ideas on why this is failing? I am trying to get mythtv installed and it depends on QT5. Any ideas would be fantastic! Thanks again for the great work.
from
steveZ
SteveZ, make sure you installed the right qt5 version for -curent and its dependencies (only libxkbcommon on -current, AFAIK):
32 bits: https://slackware.nl/people/alien/sbrepos/current/x86/qt5/
64 bits: https://slackware.nl/people/alien/sbrepos/current/x86_64/qt5/
Cheers.
Steven, it looks (from the string “GLIBCXX_3.4.26”) that perhaps you have not upgraded to the latest gcc packages.
And even if you do not have gcc installed, you should have upgraded to the latest aaa_elflibs package for -current.
Also, the list of external dependencies for qt5 on -current is:
SDL_sound, OpenAL, libxkbcommon.
Dude, your Chromium slackbuild scripts are the best.
Thank you so much for taking the tame to get all that together and sharing it, so bad ass awesome.
Regards,
I have been testing dxvk with wine and found it doesn’t work in Slackware.
The problem is that I am missing 32-bit versions of the vulkan icd files in ‘/usr/share/vulkan/icd.d/’.
After manually adding ‘/usr/share/vulkan/icd.d/radeon_icd.i586.json’ and creating a compat32 package for vulkan-sdk I found that I can not compile 32-bit wine using ‘–with-vulkan’ and can run a game that requires 32-bit vulkan working.
Can this please be fixed in your multilib repo so that in the future dxvk can be supported at SBo? Note this will probably not be until 15.0 due to lacking c++ support in 14.2, but current users could benefit in the meantime.
orbea, you write “I found that I can *not* compile 32-bit wine using ‘–with-vulkan’ “… do you in fact mean the opposite “I found that I *can* compile 32-bit wine using ‘–with-vulkan’ “?
And how ignorant of the vulkan developers to add architecture-dependent files in /usr/share/ .
But if your answer to my first question is “yes, the *not* needs to go” then I can update the convertpkg-compat32 script to make an exception for the vullkan-sdk package.
alienbob, Sorry, that is a typo where ‘not’ should be ‘now’. I meant to write, ” found that I can now compile 32-bit wine using ‘–with-vulkan’ “.
Also the vulkan icd files are in the mesa package, not in the vulkan-sdk package. The vulkan-sdk is needed during compile time and the icd files in mesa are needed at run time. I agree about not adding architecture independent files to /usr/share, but when I asked in #dri-devel @ freenode I was told they are following the XDG standard where loader searches are in XDG_DATA_DIRS.
I was also linked this old commit for intel which explains some of the reasoning.
https://gitlab.freedesktop.org/mesa/mesa/commit/a2ae67aa47df6dd41802462105dac7ce5ab676a2
orbea, the updated compat32-tools and *-compat32 packages that I just uploaded should fix your mesa issue.
Thank you, that worked. I also failed to notice you already have a compat32 vulkan-sdk package and forgot to mention that a quick test on my system is whether /usr/bin/32/vkcube works or not.
Installed Slackware Current and KDE5 on a new partition. After much wailing, gnashing of teeth and rending of garments, I finally got Firefox and Chromium to stop crashing upon startup. I couldn’t get a dvd install to work because the kernel version was too old to recognize the new AMD SATA controller, which left me in initrd hell. I rsync’d the packages from mirrors.kernel.org using a fedora30 install on a usb drive, chrooted into the slackware partition after manually extracting the A packages and running install/doinst.sh as necessary, then using installpkg thereafter. Things were a bit unstable until I added the following to /etc/fstab:
tmpfs /dev/shm tmpfs defaults,nodev,nosuid,noexec,size=4g 0 0
That finally settled things down allowing me to show up here and pass on the required fstab entry.
Thank you for all of your work creating the awesome Slackware and Slackbuild packages that just work. Pat must be proud.
Henry, the Slackware installer adds that ‘tmpfs’ line to your computer’s /etc/fstab – modern browsers need it, and not having it may well be the cause of all your system’s instability.
You say that the DVD install did not work because your SATA controller was not recognized, yet your computer seems to be running Slackware-current now. So I must assume you used a Slackware 14.2 DVD instead of downloading a slackware-current ISO and using that? Booting a Slackware-current installer should work.
Another hint for the future (and other people visiting here): instead of putting a Fedora installer on a USB stick, you can also download the small XFCE ISO of Slackware Live. It is based on slackware-current and it will show you quickly whether Slackware-current will recognize your hardware and work out of the box.
But, I am glad you made it work in the end. You may have learnt a lot in the process 😉
Primary lesson: don’t resize a mounted partition. I wanted to install archlinux to try wayland, since /usr/bin/startplasmacompositor hangs with “no backend specified..trying auto resolution.” In my slumber I resized the Slackware current that I’ve been enjoying for months. Didn’t notice the error in my ways until the next day when boot couldn’t mount root and e2fsck found bazillions of block errors. In hindight, I should have used current usbboot.img on a thumb drive, which is about the safest way to install Slackware, had I been awake.
Should this episode show up in a search engine, heed Master Hameleers’ advice.
Dolphin has been generating these messages when launched from konsole. (version 19.04.2 and previous)
kf5.kio.core: We got some errors while running testparm “Load smb config files from /etc/samba/smb.conf\nError loading services.\n”
kf5.kio.core: We got some errors while running ‘net usershare info’
kf5.kio.core: “Can’t load /etc/samba/smb.conf – run testparm to debug it\n”
I traced the issue to /etc/kde/xdg/kio.categories:kf5.kio.core.sambashare. I ripped out samba; the messages vanished.
This message remains a mystery to me. Just a benign message, doesn’t show up in /var/log/*.
kf5.kio.core: “Could not enter folder tags:/.”
Henry, those messages are caused by configuration file/directory that are absent on your computer. It’s normal – these are not created by the OS but by the system administrator, only if they are needed. So ignore them.
And next time, post your comment below the relevant article will you? This should not have been posted in the generic “your feedback” page, this does not provide the required context for other visitors.
Sorry for the non sequitur. I only noticed the messages when launching applications from konsole commandilne when too lazy to reach for the mouse. Had no idea where they came from or where to post. I’ll go away now…
Hey, Henry. you must be new!
When booting Slackware64-14.2 Live 1.2.0, I could put Ctrl where it belongs (that is, where it was almost universally from 1963 to 1986) in virtual consoles and X with these boot parameters:
kbd=emacs2 xkb=us,,ctrl:swapcaps
However, the layout was not carried over when I ran setup2hd. Typing those parameters on the “LILO append” line in setup2hd didn’t work, either.
In the Live OS, these files seemed to hold my settings:
/etc/rc.d/rc.keymap
/etc/vconsole.conf
/etc/X11/xorg.conf.d/30-keyboard.conf
Manually recreating them in the installed OS did the trick.
At least the Live OS showed me the way, and I am grateful for that. Thank you.
Hi
When trying to download a package from your main server I get this message
–2019-06-21 15:47:28– https://slackware.nl/slackware/slackware64-current/slackware64/d/icecream-1.2-x86_64-1.txz
Resolving slackware.nl (slackware.nl)… 163.172.25.142
Connecting to slackware.nl (slackware.nl)|163.172.25.142|:443… connected.
ERROR: cannot verify slackware.nl’s certificate, issued by ‘CN=CAcert Class 3 Root,OU=http://www.CAcert.org,O=CAcert Inc.’:
Self-signed certificate encountered.
To connect to slackware.nl insecurely, use `–no-check-certificate’.
Also slackpkg does not update to the latest updates
Best regards
Xavier, if you go to http://slackware.nl/ you’ll see a yellow colored link that takes you to http://alien.slackbook.org/blog/adding-cacert-root-certificates-to-your-slackware/ with instructions on how to fix that warning.
Cheers!
Thanks Ricardo
Did follow your advice
cheers
Hi Eric,
I spotted the following on your resume:
“…design of a 9000-user global LDAP directory-infrastructure…”
Woof!
Could I entice you to post an article on how to plan an LDAP server? At work I want to install some kind of central authentication. An article from somebody who configured a 9000 user server would much help the RTFM ritual. 🙂
Hi Upnort.
This was 15 years ago, using Redhat Linux 8, which comes with an LDAP server plus tools out of the box, and so do all the RHEL releases that came after.
Slackware is not equipped to run an LDAP server, or central authentication. It will be really clumsy and possibly insecure, unless you add PAM.
You can easily modify the openldap-client SlackBuild in Slackware or use my ancient version of openldap-server in my own repository if you want, but setting it up is so non-trivial that a simple article will not be enough. At the time I scoured the internet for information and to get familiarized with it.
I will not write an article on how to setup an LDAP server because it is way out of scope for the things I do nowadays.
Mr Alienbob,
Your glibc multilib does not include stdatomic.h, which a bit of code I’m writing requires, is this purposeful? I would prefer this be included in this package by default.
Hi Matthew
The file stdatomic.h is part of the compiler package, it does not belong to glibc.
Either you use the one from gcc (multilib as well as the regular Slackware package) or else from clang, depending on your preference.
forgive me, I saw it in my compiler package list, but my linker does not seem to pick it up, I’ve compile the software in question on another computer, it’s a manjaro install and things do not complain there.
Addendum: I’m using the syntax #include for this, I’m presuming I would have to alter my compiler options to target the lib explicitly, but this annoys me, since the lib itself is in a version numbered dir, and would thus be non-portable.
Matthew, if you seek functional support for your programming related questions, then this blog is not the right place. Please ask in a programming related forum, or of you think your issue is Slackware-specific, create a thread in a Slackware forum such as the one on linuxquestions.org.
Heyo alienbob, after some additional searching, it appears that you were correct, it was an issue with meson, combined with some underlying dysfunction with my code-checking software, the linker finds stdatomic.h, but the checker does not. My apologies for bothering you unnecessarily.
Eric, there was a mention of you and your multilib packages in an article on cross-compiling using gcc posted at opensource.com: https://opensource.com/article/19/7/cross-compiling-gcc
Here is the relevant quote from the article with some context:
Kudos!
Hi Eric!
trying to download the newest chromium-pepperflash-plugin gives me cryptopp instead:
https://slackware.nl/people/alien/slackbuilds/chromium-pepperflash-plugin/pkg64/current/
Can you look into the issue please?
Thanks again for all your help.
Eduardo. medication is working against me. Trying to reduce my nerve pain meds because they kill my spark, but as a result I suffer from detox. Fixing the misplaced packages now.
Thank you Eric. Don’t worry. Now I could download and upgrade the pepperflash plugin. Hope your health keeps improving!
Hi Eric
Firstly thanks in that I used your slackware64-live-xfce-current to dig myself out of a hole. Earlier today I upgraded to the latest 14.2+, ktown and multilib on one of my virtualbox instances. Forgot to upgrade my initrd.gz and grub.config. Yikes!
So downloaded slackware64-live-xfce-current.iso – nice and small.
Booted slackware64-live-xfce-current.iso successfully
Mounted my harddrive SDA1
chrooted to it
Successfully created initrd.gz and grub.config
Rebooted the instance successfully. Magic
Thank you ever so much.
Possibly could have used a 14.2+ install disk, but where would the fun be in that.
Yeah it’s nice that liveslak is not just a research project but also a very useful tool 🙂
Hi, Eric,
Just a trivial point. While i don’t usually watch the current package updates, I happened to notice that LLVM is now 8.01. From llvm.org, perhaps we could change slack-desc to include:
“The LLVM Project is a collection of modular and reusable compiler and toolchain technologies. Despite its name, LLVM has little to do with traditional virtual machines. The name “LLVM” itself is not an acronym; it is the full name of the project.”
Thus omitting “Low Level Virtual Machine”?
-Hank
Hi Henry,
You should direct comments about the content of Slackware itself to Pat Volkerding rather than to me. There’s a thread on LinuxQuestions where precisely this type of remarks/requests can be posted, Pat is watching what is posted there.
Hi Eric!
I would like to know the difference between slackware-current ffmeg-4 package and yours except libraries place (external vs included).
And the regarding above when to use each, as I assume they can not coexist?
Zdenko
Of course they cannot co-exist. You install either the Slackware package of ffmpeg, or mine.
My version of the package comes with more codec and device support, and those codec/device libraries are statically compiled into ffmpeg so that my package does not have external dependencies outside of Slackware.
That’s basically it.
Hey Bob (Eric),
First and foremost: THANK YOU very much for keeping Slackware alive!!!
I’m not sure if this is possible, but are you going to move to the 5.x kernel?
That would probably mean moving forward with a new version, I’m guessing?
Maybe even a new OS SlacOS 2.0 – Sternware 1.0 – Rigidware 1.0
The current maintainer is living up to the name slacker.
I’m not trying to be mean, I just don’t want to see it disappear.
Thanks to you as well as others, the cause is still alive.
You should ask to be given co-ownership, start a Board of Slackers
Get the system out of a few peoples hands and into a community so the next generation will keep it going.
A system / organization is only as strong as it’s followers.
Well, here’s to keeping it alive!
Thanks Again,
Rufeo
Hi Rufeo
Since I am not the maintainer of Slackware Linux I will not make the call to upgrade to a 5.x kernel. I am happy with the Slackware-current 4.19 series and also use that on my Slackware 14.2 server. Patrick is of course the Original Slacker of this distro, and he takes his time getting stuff just the way he wants them.
In the meantime I am not going to fork anything, I work with Pat and the other members of the Coreteam to test and integrate new stuff. I like it that way.
Yes indeed!!!
Updated Slackware 14.2 -current today!
Kernel: 5.4.1
Web Server: Apache/2.4.41 (Unix)
Server Side: PHP 7.4.0 / Python 3.7.5
First Linux OS to update to 5.4!
Thank you!
Thank you!
Hi AlienBob
I use Slackware 14.2 with kde4 and multilib.
I’m slowly coming round and for the last few weeks use your chromium and qt5. So thank you.
I’m a heavy libreoffice user. I use macros to transform gnucash reports so I can prepare financial reports for my colleagues. I use base with a mariadb back end to maintain a subscribers list again with macros to generate various reports.
For a long time I’ve used libreoffice, from SlackBuild, maintained by Willy Sudiarto Raharjo. It’s worked for me, though there have been times when it’s been a bit buggy – not because of anything Willy has done, but because of regressions introduced at source by the developers.
I understand, though I be could be wrong, that libreoffice no longer supports kde4 integration.
Today, I’ve compared the Willy Sudiarto Raharjo package with the package maintained by your self.
With both packages I’m exporting SAL_USE_VCLPLUGIN=gtk in /etc/profile.d/libreoffice.sh
In the Willy Sudiarto Raharjo package I’ve lost the kde save dialog and also I get the gtk icons in the task bar instead of the writer, calc and base incons.
In the alienbob package all I’ve lost is the kde save dialog.
It would be nice to get the kde save dialog back, but I doubt if that is possible if what I said above that “libreoffice no longer supports kde4 integration” is true.
Are you able to advise?
Alex
Hi Alex
I have searched online forums and found discussions going back as far as 2012 about the issues with “kde-integration”. Always the advice was to use GTK UI support instead of KDE4 UI support to get working menus and dialogs in Libre Office.
Therefore I will no longer ship a “kde-integration” package in future, and instead add Qt5/Plasma5 support.
Hi alienbob,
I have found a workaround for building with clang and -m32 while using your multilib llvm package by symlinking /usr/lib to /usr/lib32. This allows building a 32-bit version of wine-4.14.
This old upstream issue seems related.
https://bugs.llvm.org/show_bug.cgi?id=12108
And I have been told gentoo is unaffected, but I can not figure out why from their documentation or ebuilds.
https://gitweb.gentoo.org/repo/gentoo.git/tree/sys-devel/llvm
https://wiki.gentoo.org/wiki/Project:AMD64/Multilib_layout#Why_lib_and_not_lib32.3F
https://wiki.gentoo.org/wiki/Project:LLVM/Multilib
Maybe you could consider adding a /usr/lib32 symlink to your multilib packages to work around this for now? Or do you have a better idea of why this is happening and how to solve it?
I made a more detailed explanation of what I have found in this old Slackware thread which is related, but still unsolved.
https://www.linuxquestions.org/questions/slackware-14/compiling-32bit-llvm-on-multilib-4175602559/#post6030352
Thanks for gcc and soffice, on your vacation no less!
Just FYI. Just installed the latest VLC (restricted version), thank you very much, but it returns the following message,
“bash-5.0$ vlc
VLC media player 3.0.8 Vetinari (revision 3.0.8-0-gf350b6b5a7)
Segmentation fault
bash-5.0$”
I’m running a fairly recent installation of -current along with your multilib files.
Just re-installed VLC-3.0.7 and it is working perfectly.
Thanks for all your hard work.
CWizard, runs fine here on the laptop with 64bit Slackware-current and multilib, and Plasma5.
The external runtime dependencies are qt5, libxkbcommon, OpenAL, SDL_sound. What versions of these packages do you have installed?
Can you run “strace vlc” to see if any error shows up that you’d recognize? Does the ncurses version “nvlc” work in a terminal?
Here you go. As requested.
🙂
Hope it helps.
Thanks, again.
> What versions of these packages do you have installed?
qt5- 5.12.3
libxkbcommon-0.8.1
OpenAL-1.9.1
SDL_sound-1.0.3
> Does the ncurses version “nvlc” work in a terminal?
No the same result, a Segmentation fault.
> Can you run “strace vlc” to see if any error shows up that you’d recognize?
access(“/etc/ld.so.preload”, R_OK) = -1 ENOENT (No such file or directory)
openat(AT_FDCWD, “/usr/lib64/locale/locale-archive”, O_RDONLY|O_CLOEXEC) = -1 ENOENT (No such file or directory)
There were 14 UF-8 related messages similar to this one,
openat(AT_FDCWD, “/usr/lib64/locale/en_US.UTF-8/LC_MEASUREMENT”, O_RDONLY|O_CLOEXEC) = -1 ENOENT (No such file or directory)
openat(AT_FDCWD, “/usr/lib64/tls/x86_64/x86_64/libGL.so.1”, O_RDONLY|O_CLOEXEC) = -1 ENOENT (No such file or directory)
stat(“/usr/lib64/tls/x86_64/x86_64”, 0x7ffc3e168c00) = -1 ENOENT (No such file or directory)
openat(AT_FDCWD, “/usr/lib64/tls/x86_64/libGL.so.1”, O_RDONLY|O_CLOEXEC) = -1 ENOENT (No such file or directory)
stat(“/usr/lib64/tls/x86_64”, 0x7ffc3e168c00) = -1 ENOENT (No such file or directory)
openat(AT_FDCWD, “/usr/lib64/tls/x86_64/libGL.so.1”, O_RDONLY|O_CLOEXEC) = -1 ENOENT (No such file or directory)
stat(“/usr/lib64/tls/x86_64”, 0x7ffc3e168c00) = -1 ENOENT (No such file or directory)
openat(AT_FDCWD, “/usr/lib64/tls/libGL.so.1”, O_RDONLY|O_CLOEXEC) = -1 ENOENT (No such file or directory)
stat(“/usr/lib64/tls”, 0x7ffc3e168c00) = -1 ENOENT (No such file or directory)
openat(AT_FDCWD, “/usr/lib64/x86_64/x86_64/libGL.so.1”, O_RDONLY|O_CLOEXEC) = -1 ENOENT (No such file or directory)
stat(“/usr/lib64/x86_64/x86_64”, 0x7ffc3e168c00) = -1 ENOENT (No such file or directory)
openat(AT_FDCWD, “/usr/lib64/x86_64/libGL.so.1”, O_RDONLY|O_CLOEXEC) = -1 ENOENT (No such file or directory)
stat(“/usr/lib64/x86_64”, 0x7ffc3e168c00) = -1 ENOENT (No such file or directory)
openat(AT_FDCWD, “/usr/lib64/x86_64/libGL.so.1”, O_RDONLY|O_CLOEXEC) = -1 ENOENT (No such file or directory)
stat(“/usr/lib64/x86_64”, 0x7ffc3e168c00) = -1 ENOENT (No such file or directory)
6 like similar to this,
openat(AT_FDCWD, “/usr/share/locale/en_US.UTF-8/LC_MESSAGES/libc.mo”, O_RDONLY) = -1 ENOENT (No such file or directory)
ioctl(2, TCGETS, {B38400 opost isig icanon echo …}) = 0
openat(AT_FDCWD, “/usr/share/vlc/../locale/en_US.UTF-8/LC_MESSAGES/vlc.mo”, O_RDONLY) = -1 ENOENT (No such file or directory)
openat(AT_FDCWD, “/usr/share/vlc/../locale/en_US.utf8/LC_MESSAGES/vlc.mo”, O_RDONLY) = -1 ENOENT (No such file or directory)
openat(AT_FDCWD, “/usr/share/vlc/../locale/en_US/LC_MESSAGES/vlc.mo”, O_RDONLY) = -1 ENOENT (No such file or directory)
openat(AT_FDCWD, “/usr/share/vlc/../locale/en.UTF-8/LC_MESSAGES/vlc.mo”, O_RDONLY) = -1 ENOENT (No such file or directory)
openat(AT_FDCWD, “/usr/share/vlc/../locale/en.utf8/LC_MESSAGES/vlc.mo”, O_RDONLY) = -1 ENOENT (No such file or directory)
openat(AT_FDCWD, “/usr/share/vlc/../locale/en/LC_MESSAGES/vlc.mo”, O_RDONLY) = -1 ENOENT (No such file or directory)
And, finally,
openat(AT_FDCWD, “/usr/lib64/libgraphite2.so.3”, O_RDONLY|O_CLOEXEC) = 3
read(3, “\177ELF\2\1\1\0\0\0\0\0\0\0\0\0\3\0>\0\1\0\0\0\2401\0\0\0\0\0\0″…, 832) = 832
fstat(3, {st_mode=S_IFREG|0755, st_size=153712, …}) = 0
mmap(NULL, 155864, PROT_READ, MAP_PRIVATE|MAP_DENYWRITE, 3, 0) = 0x7f441c042000
mmap(0x7f441c045000, 114688, PROT_READ|PROT_EXEC, MAP_PRIVATE|MAP_FIXED|MAP_DENYWRITE, 3, 0x3000) = 0x7f441c045000
mmap(0x7f441c061000, 20480, PROT_READ, MAP_PRIVATE|MAP_FIXED|MAP_DENYWRITE, 3, 0x1f000) = 0x7f441c061000
mmap(0x7f441c066000, 12288, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_FIXED|MAP_DENYWRITE, 3, 0x23000) = 0x7f441c066000
close(3) = 0
mprotect(0x7f441c066000, 8192, PROT_READ) = 0
mprotect(0x7f441c618000, 4096, PROT_READ) = 0
mprotect(0x7f441c1a5000, 4096, PROT_READ) = 0
mprotect(0x7f441c192000, 4096, PROT_READ) = 0
mprotect(0x7f441c6ce000, 28672, PROT_READ) = 0
mprotect(0x7f441c15c000, 4096, PROT_READ) = 0
mprotect(0x7f441c1ae000, 4096, PROT_READ) = 0
mprotect(0x7f441c1da000, 8192, PROT_READ) = 0
mprotect(0x7f441c401000, 4096, PROT_READ) = 0
mprotect(0x7f441c719000, 8192, PROT_READ) = 0
mprotect(0x7f441c879000, 32768, PROT_READ) = 0
mprotect(0x7f441c93e000, 12288, PROT_READ) = 0
— SIGSEGV {si_signo=SIGSEGV, si_code=SEGV_MAPERR, si_addr=0x2270} —
+++ killed by SIGSEGV +++
Segmentation fault
bash-5.0$
My VLC packages were compiled against my qt5- 5.13.0, libxkbcommon-0.8.4 and OpenAL-1.19.1 packages on Slackware-current. I notice you have older versions of all of those. How did you obtain/build/install these packages?
The ‘strace’ output did not help me find a clue unfortunately.
Updating qt5-5.12.3 to qt5- 5.13.0 did the trick.
Thanks, again.
Hey Bob (Eric),
First and foremost: THANK YOU very much for keeping Slackware up to date!!!
I have a questions.
I have been using the 14.2 -current for a while and have conquered many obstacles Slackware presents: https://www.team3.org/index.php?ps=a
You update it quite regularly (YAY!), one thing that bugs me is the deleting of each individual file and then their directories. Wouldn’t “rm -rf /path/to/the/base/directory” work much faster?
I know that deleting each file and folder is part of a log, but that could be produced quicker with “ls /path/to/the/base/directory > deletelog.log”
I am giving basic concepts, I’m sure there a better and faster ways of doing things, I’m just hoping for a possible way of speed up the delete process.
Also, I use a variety of flash based devices (SSD, CF & USB) and the deleting of each files and directory every time I update Slackware causes me to cringe… as we know deleting causes the most wear on flash technology. Slackware does not seem to SSD friendly.
Wear & tear are my real concerns, speeding up the process would be nice…
Thanks Again,
Rufeo
Hi Rufeo,
I assume you are talking about the Slackware package tools, like “upgradepkg”? I do not maintain Slackware and the feedback about the distro should go to the distro, not to me. There’s a topic on linuxquestions.org that deals with all the suggestions people have for improving the distro: https://www.linuxquestions.org/questions/slackware-14/requests-for-current-14-2-15-0-a-4175620463/
What you describe though, is standard practice for people running Slackware-current. Lots of packages get upgraded all the time, and yes… that means a lot of files are going to be deleted when you upgrade the next batch.
If you do not like that, you should revert to the stable version 14.2 instead of taking part in our Beta testing.
I use:
slackpkg update
slackpkg install-new
slackpkg upgrade-all
grub-mkconfig -o /boot/grub/grub.cfg
slackpkg clean-system
Thanks,
Rufeo
Thanks a lot for the swift delivery of glibc-multilib upgrades! Especially since I guess you have quite a backlog both at work and with your repos after your vacation,
Just tried building the new chromium version using the build infrastructure from you repository. In the SlackBuild I just changed the version to 77.0.3865.75. It downloaded the source and started processing. After about three hours or so the build failed on a seeming inconsequential error –
[37814/37815] AR obj/headless/libheadless_shell_lib.a
[37815/37815] LINK ./chrome
— Manually installing everything.
install: cannot stat ‘chrome/app/theme/chromium/product_logo_22.png’: No such file or directory
chromium.SlackBuild FAILED at line 1079
Is this recoverable? If not where does this .png file need to be. Thanks in advance.
Updating Chrome was on my TODO list after releasing new Plasma5 packages. I already have the SlackBuild ready, and this is what you need to apply:
Next time though, please post your comment about chromium in the most recent chromium related article instead of the generic Feedback page.
OK, thanks Eric.
Hi Eric,
I have an old Epson V330 flatbed scanner which I seldom use. Yesterday I tried scanning a few pages but couldn’t get past the message from iscan: “Could not send command to scanner. Check the scanner’s status.” The scanner showed up in lsusb, and it worked perfectly when I tried vuescan (the free test version with watermarks).
Don’t know if this could be an effect of your iscan and iscan-data upgrades from 17 September? Or maybe some other -current upgrade, e.g. the sane upgrade from 1 August? Have you encountered any issues yourself?
I know that Epson scanners are not the easiest to get working in Linux, and I have workarounds so there’s no panic.
I do not use iscan myself and upgraded the packages because someone asked me to here on the blog. The iscan program had stopped working as a result of (I think) the eudev upgrade.
So there is nothing I can offer to assist, hopefully someone else who uses iscan has an idea.
Iscan continued: I downloaded the package https://download2.ebz.epson.net/iscan/plugin/perfection-v330/rpm/x64/iscan-perfection-v330-bundle-2.30.4.x64.rpm.tar.gz, extracted and created new iscan, iscan-data, and esci-interpreter-perfection.-v330 packages using rpm2tgz. Upgrading to fresh esci-interpreter package made no difference. I then replaced your iscan-data and iscan-data packages with the ones from the bundle – and that did the trick, my scanner is now recognised and works.. So something seems to dodgy with your packages – at least on my machine, but of course that might have to do with other stuff installed.
One difference I notice (but I don’t know if it is crucial) is that the package I created from rpm contains /var/lib/iscan, which is not part of yours.
Well, I will wait for someone to debug my package and tell me what to fix. Without the Epson hardware, there is not much I can do.
Hello!
Do you have any packages that install Qemu + Virt-manager on a Slackware64 machine installed? I would like to do some experiments, but there are a lot of dependencies and also the modules that will be installed in the kernel, as well as the necessary modifications to libvirt …
I’m afraid of trying to install everything manually with the Slackbuilds packages and damage my system by some error.
Thanks for your help!
Hi Eric.
I’m working on Slackware current, and i have a problem with Aegisub. When I am trying to install this I have a problem with wxWidgets. I have message “configure: error: wxWidgets detection failed, please set –with-wx* or add the libraries to your LIBS, CXX/CFLAGS.” Do You have any idea what do whit this?
Thanks for help
I have no idea, Aegisub is not part of my repository.
@Arkadiuz: do you have all dependencie for Aegisub installed (e.g. wxGTK3)? See https://slackbuilds.org/repository/14.2/multimedia/aegisub/
Thank you! 🙂
Hi Eric,
I’m trying to boot multiple computers into liveslak using pxe. And everything works perfectly! …as long as I specify a nic= statement boot param with a good driver. If I do not then the boot process stops while trying to mount nfsroot because network device is not found. Once I am dumped to a command line I can modprobe forcedeth and then the network shows up in /proc/net/dev and `ip a` shows I get a dhcp ip address just fine.
I am trying to diagnose the issue looking at liveinit.tpl. At line 413 in setnet() your comment indicates a network driver should have already been loaded if we haven’t specified one in our boot options. So I reach line 454 “assuming ‘eth0’. Trouble ahead.” with no network driver loaded presumably because it failed to autodetect a device prior to ever entering setnet()? Am I right to think this is a problem with the normal init process and not the liveslak specific part? Some digging shows that line 294 should be finding network drivers, but all udevadm returns for me is the loopback device, suggesting I have a problem even earlier than that.
I am getting my information on what “should” happen from https://docs.slackware.com/slackware:liveslak – which specifically says “If you are on a network where a DHCP server configures your clients, then the nic parameter will not be needed as Slackware Live Edition will figure out all the details by itself.” – As well as not specifying any nic= line, using nic=auto:eth0:dhcp doese not work, presumably because “auto” is not a valid driver to modprobe.
I see absolutely no warnings or errors in the boot process prior to the line 454 error. Any advice or speculation is appreciated.
Hi Kanaya,
You can try adding “debug=4” to the boot commandline. What will happen is:
1- the init script will execute verbosely
2- the udevadm command will be verbose
3- the boot process will drop you into a console *right* before exiting the initrd. You’ll be able to inspect the boot logs (hopefully they did not scroll out of the buffer) and when you want to boot further, simply type “exit” to exit the initrd.
First off thanks for your “debug=” advice – it proved very useful in determining which branches are running (and therefore failing to be useful). After roaming through numerous `modinfo` and `udevadm info` and comparing differences to working machines, I’ve come to some conclusions.
Network driver doesn’t load on VM because you don’t have virtio_pci.ko. Virtio-pci is used for “legacy” QEMU network devices (NOT “deprecated”!!!), and virtio-net is used for “modern” ones. I recommend including virtio_pci.ko anywhere you include virtio_net.ko so that QEMU is supported in any configuration that uses a virtual network device. But even if you do include virtio-pci in your initrd…
QEMU’s legacy network device isn’t a “net” device by udev’s standards… it’s a “pci” device with network capabilities. The same is true for my physical Nvidia MCP51 ethernet device. Neither of these kernel modules will get loaded by udev according to the liveinit.tpl script because in there, udev uses “–subsystem-match=net” to look for [modern] network devices.
That means on line 294 udev fails to find any net device besides loopback, and runs “udevadm trigger” who’s default action is “change” NOT “add”(!!). This whole mess is fixed by running `/sbin/udevadm –subsystem-match=pci –action=add` if we fail to find any net devices on 294. Or, we can over-simplify the whole thing: once we have $NFSROOT defined just skip any other checking and just load all included drivers with `/sbin/udevadm –action=add`, letting them sort out on their own which is needed! I’m not super familiar with initrd structures but my assumption is that this won’t cause any harm, it will only make booting take a couple milliseconds longer while parsing a bunch of modaliases. Or we can run two udevadm commands, one to add net devices, the second to load pci devices.. and assume all other obscure network devices are too vile to want to deal with.
My final thought is just that the whole “-n $NFSBOOT” block needs reworking to just load as many drivers as are included, to ensure all pxe boots are supported… but I could be missing something really obvious. What do you think? Thoughts from someone more experienced than I would be very appreciated. I hope you find this information interesting and not annoying.
P.S. Here is some information that’s not necessary, but interesting and marginally related:
-You can see how Red Hat distributes their QEMU device IDs at https://github.com/qemu/qemu/blob/master/docs/specs/pci-ids.txt – it’s unlikely that legacy devices will be deprecated anytime soon because VMs rely on consistency and longevity, even if some setups are “discouraged”.
-Line 312 in liveinit.tpl always fails; it runs `ls` looking for kernel modules in /lib/modules/$(uname -r)/*.*o, but the wildcard doesn’t glob directories so it ALWAYS loads zero modules. I’m completely bedazzled by this use and can’t figure out if it’s deliberate (to only load files you manually put there?) or if its a massive oversight that no one’s encountered or bothered to point out. It normally doesn’t run anyways, because load_kernal_modules exists and is run instead to manually load kernal modules.
Kanaya thanks for your debugging and observations, they are most helpful.
The idea was to have ‘udevadm’ first look for ‘net’ subsystem devices and if it only finds ‘lo’, then it will repeat the trigger in blanket fashion… but you correctly pointed out that I had omitted ‘–action=add’. I will add that to line 295 (which now reads “/sbin/udevadm trigger $DEBUGV”) which should solve your issue because then udevadm will enable any hardware it finds.
Your final comment about the block of code which loads the kernel modules, these lines are taken from the Slackware init script verbatim (since the Slackware ‘init’ script is what ‘liveinit’ is based on). The section “else # load modules (if any) in order” goes all the way back to old kernels with *.o instead of *.ko modules in a single directory. The .ko module extension was introduced in the Linux kernel 2.6 (i.e. in 2003) and I think it is safe to assume that Slackware Live will never be based on Slackware 11 or older… so I can clean up that bit too.
It helped to have this conversation on the blog instead of on Twitter… the above would have been lost to everybody else after a few hours.
Had an opportunity to test today and I’m tapping in to say 1.3.3 works great! Everything runs as expected on both physical and virtual machine. Now I get to break into the fun part of building custom package lists… 🙂 To be honest, I expected early on you’d just tell me I was doing something wrong and this would go nowhere, so I’m happy to see that I could be helpful improving network booting. Thank you so much for being open to the discussion.
I’ll report back if I encounter any further oddities.
Check out the examples on how to customize an ISO in https://git.slackware.nl/liveslak/tree/make_slackware_live.conf , the documentation is here: https://docs.slackware.com/slackware:liveslak#using_the_customization_feature_of_the_live_os
Let me know if you run into questions or issues.
> To be honest, I expected early on you’d just tell me I was doing something wrong
> and this would go nowhere
I only react that way to people who complain, don’t back their bug report with details and/or simply expect me to fix their issue without bothering them further.
Anyone else, I will be happy to help 😉
Running Slack4-current with latest Plasma5
Just noticed that there is a new Digikam version for Plasma5, 6.4.0. Tried building it using your infrastructure (has always worked in the past just fine) but apparently the build requirements have changed. First I received an error that x265 wasn’t available so I built that from the SlackBuild. Solved that error but now I get –
CMake Error at core/CMakeLists.txt:251 (find_package):
By not providing “Findexiv2.cmake” in CMAKE_MODULE_PATH this project has
asked CMake to find a package configuration file provided by “exiv2”, but
CMake did not find one.
Could not find a package configuration file provided by “exiv2” with any of
the following names:
exiv2Config.cmake
exiv2-config.cmake
I have exiv2 installed from Slackware but it appears (to me at least) that the build script needs some more things included. Can you point me in the right direction please? Thanks in advance
You’ll have to find the cause and the fix yourself.
I will update digikam on the next batch somewhere in November probably. I do not have the time to spend on debugging before then, sorry.
Believe it has to do with the version of exiv2. Slack64-current is at 0.26 and the new Digikam requires at least 0.27.0 (current release is 0.27.2). Downloaded source and am trying to get the SlackBuild to run with it but there are problems.
config.status: creating ./Makefile
config.status: creating ./config/config.mk
config.status: creating ./config/exiv2.pc
config.status: error: cannot find input file: `./po/Makefile.in.in’
Will keep trying to get it to build. This is an issue for Slackware and not Plasma5.
When compiling ‘mesa-19.3.0-rc1’ in 32-bit mode on ‘slackware64-current’, I get ‘/usr/include/libunwind.h:23:11: fatal error: libunwind-x86.h: No such file or directory’. Your ‘convertpkg-compat32’ script removes ‘/usr/include’ directory with ‘libunwind-x86.h’ file from ‘libunwind-1.3.1-i586-2’ when converting, and x86_64 version contains only ‘libunwind-x86_64.h’.
Hi Golem
Thanks for the report. It’s easy to fix, and the next batch of updates for the multilib repository will address this.
In the meantime if you do not want to wait you can download and apply the patch yourself:
Hello
I dont really know what is the command to install a software package such as eclipse on liveslack . i dont have problem with installing oftware in slackware but commands are not eorking in live .
Hi Maj,
I don’t really understand your problem. You need to be root to install software, that is for sure. And a Slackware Live system needs to be installed on a USB medium and persistence has to be enabled if you want to make changes to the system (such as: installing software) that you want to keep across reboots.
A Slackware Live that runs straight off the ISO image is not persistent, everything happens in RAM and all changes are lost when you reboot.
There should not be a difference between a ‘normal’ Slackware system and a Slackware Live USB stick with persistent storage enabled.
Hi,
let me thank you for your great packages first! I really appreciate the hard work you put into this! 🙂
I just found out that the lensfun package within ktown is slightly outdated, the included version is 0.3.2 which was released in 2015, the current version is 0.3.95 from 2018 (which is the newest version as this package is sadly quite understaffed with developers)
Hi Clerian, I will add a newer lensfun package on my TODO list for next ‘ktown’ release.
Hi. there is a new version of veracrypt out. Would it be possible to add it to your todo list?
Thank you!
Hi Eric, tried building the new version of Chromium yesterday but after a while it dies citing missing files. Your script has worked fine for the last couple of releases. My build directory got purged but I can kick it off again if you need the specifics. Thanks
Hi Tom
I actually have a Chromium 79.0.3945.79 package here at home. The SlackBuild needed some work. Unfortunately I am unable to get this to compile on 32bit Slackware, and the chromium packager community (https://groups.google.com/a/chromium.org/forum/#!forum/chromium-packagers) remains silent. So I am stuck and not releasing this version of Chromium for the moment. Or maybe I’ll stop with supporting 32bit and be done with it. I am undecided.
Here’s you’ll find the work that was needed to make a working Chromium 79 64bit package: https://git.slackware.nl/asb/commit/?id=dde86b8f3578675d62527804cf2d5d39d402e310
Hi Eric
FWIW………..
ArchLinux32 and Debian have been able to issue 32-bit versions of Chromium-79……
http://pool.mirror.archlinux32.org/pentium4/testing/chromium-79.0.3945.79-1.0-pentium4.pkg.tar.xz
http://ftp.us.debian.org/debian/pool/main/c/chromium/
Cheers
PeeBee
Well, both distros compile with a toolchain that is not as ancient as what’s available in Slackware 14.2. For instance, they use a recent GCC so that they do not have to compile clang from scratch and do not run into the bootstrapping issues I get.
Hi Eric,
I just updated slackware64-current, and had to use libreoffice (your buid for libreoffice), but … it says it needs lboost 1.71.0, and slackware64-current has now lboost 1.72.0. I installed back lboost 1.71.0, but I think it can not be a long term solution. Can you help me?
Thanks.
Please install my boost-compat package. It has been mentioned here and on LinuxQuestions.org quite often as a means to prevent breakage with 3rd party software when slackware-current gets updated.
Hi Eric. I use Slackware 64 current version, I upgrade once a week using slackpkg. One of the repo is yours and my blacklist contains only kernel and kernel-sources (I compile my kernel by hand). 2 days ago I’ve noticed I can’t mount any usb pendrive or my Kindle as normal user, but as root everything worked. And the problem was polkit. Last upgraded has downloaded 0.116, I removed it and reinstall 0.113 and works again. Do you know why? Thanks for your enormous job. Regards
Hi Gerardo,
Make sure to install the new polkit dependency mozjs60, slackpkg –install-new should take care of that.
Cheers!
Ok, thanks for your help !!
I’ll try .
Hello Erik I just recently installed slackware64-current multilib and there are missing 32dev.sh 32dev.zsh in /etc/profile.d/
Also I have lenovo ideapad 700 15ISK 80RU and after 15-20 min of work freezes dysplay and after short time freezes all system on original intel drivers kernel 5.4.6
No complains just mention the situation
The file /etc/profile.d/32dev.sh and /etc/profile.d/32dev.csh are part of the package “compat32-tools” which you possibly forgot to install. There is no .zsh version of this script.
As for the freezes of your graphical desktop, this blog is not the place to discuss issues like that. There’s the Slackware forum on the https://linuxquestions.org/ site where you can find a lot of good people who are willing to help you with your question.
Also, I do not take kindly to people who feel the need to comment on my own website and then manage to spell my name wrong.
Thanks for the new Handbrake package!
Hi Eric.
Thank you for the great work )
After update python 3.8 in Slackware, you didn’t recompile all needed packages.
Need recompile cantor-19.12.0-i586-1alien.
ldd report:
ldd /usr/bin/cantor_python3server not found lib(s): libpython3.7m.so.1.0
OR rebuild all KDE Application with new version released January 09, 2020.
https://kde.org/announcements/releases/19.12.1/
Thanks)
Please comment below a relevant blog post. If you had opened my latest plasma5 post you would have seen that someone else already mentioned cantor.
If you use cantor, you can recompile it yourself . Otherwise just wait until my next monthly release.
Hello everyone! I know this is rather lame, but I never use Linux and I really want to learn. Is Slackware a bad choice for a first distribution? I have zero CLI experience, which I am OK with taking on the challenge. There’s just something about Slackware that makes me smile, and I’m not sure what it is but I am determined to learn it :-). I’m not afraid to learn and read manpages, but I can’t seem to understand the updating process for Slackware. I made sure to uncomment a mirror for updating in the /etc/slackpkg/mirror file (http). I then tried the following:
# slackpkg update
# slackpkg install-new
# slackpkg upgrade-all
# slackpkg clean-system
Then after I do all of that my manpages are gone! lol……ugh. Maybe this is way deeper than I understand. Lmao oh well, so on to my point. Should I even be updating anything? The bottom line is, I want to learn python 3 (latest) and maybe have the newest mozilla browser and a few other necessary things that need current versions. This is also helpful to note I am in VirtualBox, so that may be my problem. I also want to thank everyone for all of the amazing work you do for the world.
PS. Oh yea also, when I do a clean install I don’t get the option for the hugesmp.s kernel. I have an i7 (8 core), not sure if that even matters. OK! Thanks again everyone, I will be up late figuring this out 🙂
Hi Dimitri,
The set of commands you show are correct, nothing wrong with that. Your post lacks some details that are required if anyone is going to help you.
What version and architecture of Slackware did you install? Is this Slackware 14.2 64-bit? And what exact line did you un-comment in /etc/slackpkg/mirrors ?
If you browse to https://www.linuxquestions.org/questions/slackware-14/ and create an account there on LQ, you will find that Slackware forum to be full of knowledgeable people who are willing to answer all your questions during your journey to learn all about Slackware.
Thanks Eric! I won’t disappoint! I think part of the journey is figuring it out.
Hi Eric, i have been advised on linuxquestions to post my problem here as well. I have been using slackware 14.2 for couple of years now, never had any problems (that i didn’t cause) but recently got a new laptop (AMD Ryzen 5 3500u) and decided to try slackware64-current. I did clean install yesterday and installed kde5. As it is required, i installed openjdk too (got it from here https://slackware.nl/people/alien/sl…pkg64/current/) using upgradepkg and i did the same with libreoffice (http://www.slackware.com/~alien/slac…/pkg64/current). However upon running libreoffice command in terminal (as user) my KDE froze and terminal for short period of time displayed these two lines:
bash-5.0$ libreoffice
mesa: for the –simplifycfg-sink-common option: may only occur zero or one times!
mesa: for the –global-isel-abort option: may only occur zero or one times!
That’s it after this i have to force shut down my laptop. Also i tried libreoffice5.4 (https://slackware.pkgs.org/current/s…lonly.txz.html) but the result was the same, only this time lines it displayed upon running are:
bash-5.0$ libreoffice5.4
javaldx: Could not find a Java Runtime Enviroment!
Warning: failed to read path from javaldx
Can you help me? LibreOffice and openjdk installation went without issues. I googled a lot and couldn’t find solution.
Hi Pavle,
The last error you show (javaldx: Could not find a Java Runtime Environment) is most likely caused by an outdated ~/.config/libreoffice/4/user/config/javasettings_Linux_X86_64.xml file which holds your Java Runtime confguration. It may point to the Java which used to be part of the GCC compiler package suite but was removed in gcc-7.
You can either just remove that file, or else go to “Tools > Options > Advanced” and explicitly select the Java 8 you’ll find there.
The crash/freeze bug you describe looks like an issue with Mesa (perhaps in combination with a Ryzen CPU) as described here: https://gitlab.freedesktop.org/mesa/mesa/issues/1421
All I can advise you is to re-check that you have the most recent versions of all the Slackware-current packages on your system, including the ones that got added since the release of Slackware 14.2 (such as isl and libglvnd and a lot more).
Eric, I just upgraded to the latest Plasma 5 that you released today (12.2.20) and have a couple of minor issues. First, when I open the Application Launcher (bottom left of screen) it starts fine but the font used under the row of five horizontal icons at the bottom of the launcher is illegible – just black blocks. Secondly, and this is trivial, the Emoji launcher doesn’t start with the Windows Key plus . as indicated in teh Plasma announcement.
That sounds like a font anti-aliasing problem. I am experiencing that for quite a long while now on my Nvidia GPU powered desktop and I had to disable anti-aliasing for the font size used for the menu display.
Hello, Eric
Thanks for all your work keeping your blog, repository, etc, alive. It’s an excellent resource and is much appreciated.
I’ve started transitioning to current and noted SDL2-* is now part the main tree. You’ve already already added SDL2-compat32 to multilib. Can you also add the rest, SDL2_{gfx, image, mixer, net, ttf} as well? Although Steam can bring in its own versions, several games from Humble Bundle, GOG.com, and via WINE need some or all compat32 versions. No rush; you’re compat32 tools make it easy enough. For some users, having them pulled in via slackpkg, for example, may be an easier option.
Thanks,
Ed
Hi Eric,
I’m getting the following on your Chromium package using Slackpkg+:
Package chromium-80.0.3987.122-x86_64-1alien.txz is already in cache – not downloading
ERROR – Package not installed! md5sum error!
Owen
Faulty download?
root@bear:# md5sum -c chromium-80.0.3987.122-x86_64-1alien.txz.md5
chromium-80.0.3987.122-x86_64-1alien.txz: OK
root@bear:# md5sum chromium-80.0.3987.122-x86_64-1alien.txz
55c762de802ff54de6b7271d73f07c26 chromium-80.0.3987.122-x86_64-1alien.txz
Hi Eric,
I am getting download errors from your repositories (current), too. The Libreoffice update took ages to download and when it was finally done, I got a “http://bear.alienbase.nl/mirrors/people/alien/sbrepos/current/x86_64/libreoffice/libreoffice-sdkdoc-6.4.1-noarch-1alien.txz: download error
libreoffice-sdkdoc-6.4.1-noarch-1alien.txz: md5sum” for each file.
Also, running “slackpkg update”
https://mirrors.slackware.com/slackware/slackware64-current/ktown/MANIFEST.bz2:
2020-03-03 17:09:52 ERROR 404: Not Found
2020-03-03 17:09:53 URL:http://bear.alienbase.nl/mirrors/people/alien/restricted_sbrepos/current/x86_64/MANIFEST.bz2 [31703/31703] -> “/tmp/slackpkg.IyorM0/SLACKPKGPLUS_restricted-MANIFEST.bz2” [1]
!!! N O T I C E !!!
Repository ‘alienbob’ does not contains MANIFEST.bz2
Don’t worry… it will work fine, but the command
‘slackpkg file-search’ will not work on that
repository
2020-03-03 17:11:47 URL:http://repository.slacky.eu/slackware64-14.2/MANIFEST.bz2 [86008/86008] -> “/tmp/slackpkg.IyorM0/SLACKPKGPLUS_slacky-MANIFEST.bz2” [1]
!!! N O T I C E !!!
Repository ‘ktown’ does not contains MANIFEST.bz2
Don’t worry… it will work fine, but the command
‘slackpkg file-search’ will not work on that
repository
Not sure if this is important, just don’t remember seeing this notice before.
Thanks
Try a mirror like slackware.uk instead of my own ‘bear’ server if your country does not have good connectivity to my France-based server.
I generally use the slackware.uk mirror as it is usually faster in Beijing than the others for me. I just now used it to upgrade Chromium, widevine, and pepperflash. No problems. Hope this info is helpful. My system in slackware-64-current and the latest plasma5. No multilib at present.
Hmmm…interesting, I keep getting the error : (
I’ll have to tinker
Owen
I need broadcom_sta wifi module for liveslack 1.3.5 xfce. So far I’ve only found /people/alien/liveslak/optional/0060-broadcom_sta-6.30.223.271_4.4.38-current-x86_64.sxz.
I’m feeling particularly braindead at the moment – any suggestions? I’d accept “go fly a kite”…:-)
I recently inherited a Macbook Air with broadcom 4360 wifi & no wired net. Would love to get liveslack running on it.
Cheers, Another Bob
I uploaded a module for the -current kernel 5.4.25, perhaps it helps.
You can use the SBo files for broadcom-sta on Slackware-current and that will build a good package. Then you can convert that package into a live module using the script which is part of liveslak:
# makemod -i broadcom-sta-6.30.223.271_5.4.25-x86_64-3_SBo.tgz 0060-broadcom_sta-6.30.223.271_5.4.25-current-x86_64.sxz
Thank you for quick response. Pleasant surprise.
Unfortunately, liveslack xfce has kernel 5.4.20. Not having success trying to force load 5.4.25 wl.ko.. I tried using SBo but liveslack xfce does not include /usr/src/linux-5.4.20; I downloaded that kernel source but ran into “include/generated/autoconf.h or include/config/auto.conf are missing” when executing “sudo sh broadcom_sta.SlackBuild”.
I like my Macbook Air for its battery life & traveling convenience. I always take an ancient netbook upgraded with intel wifi & ssd on which I run linux & liveslack.. Just was hoping to leave netbook at home. Oh well…:-)
This page is 2.54 megabytes heavy. 🙁
Takes several minutes to load on my shitty internet connection.
=======
Stupid question…
Eric, how do I build gcc-multilib from your multilib slackbuilds?
Run fetch-from-svn-and-prep-tarball.sh from the main -source tree, and then copy your gcc-multilib.SlackBuild isl-0.18.tar.bz2 gcc.strsignal.diff into the original source tree and run ./gcc-multilib.SlackBuild?
I take the Slackware sources for gcc (non-multilib), add my own gcc-multilib.SlackBuild to them, and then run my script. This requires a multilib version of the gcc packages to be installed already.
If you want to build another version or a git checkout, your mileage may vary. You’ll have to do all the checks & validations yourself that I normally do for you.