Your feedback


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

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.

My old feedback page got too big with all the comments, so I archived it here: “your older feedback

Continue to my blog articles.


86 thoughts on “Your feedback

        1. Aegisub should compile against wxGTK3 if you add this parameter to the configure (assuming you installed my wxGTK3 package):

          # configure –with-wx-config=/usr/bin/wx-config-gtk3

  1. Hi there!

    Are you aware of any problem with Plasma starting in the last couple of days? I am not sure when exactly.

    Today I had to log out and when tried to log in again I simply couldn’t. I started to troubleshoot the issue and initially there was apparently a problem with “baloo” that was generating tons of log messages (similar issue than explained here, after stopping baloo I was able to log in again, but I had to try to reproduce the problem … so I logged out and in again, and I just couldn’t. I tried to identify the problem but unsuccessfully so far. So I just removed the whole KDE and installed again from your repository, but with no luck. I can open other desktop environments, as XFCE that I am using right now without issues, and will continue troubleshooting to see if I can find the problem, but just wonder if you (or any of your followers) is experiencing similar problems.


    1. I am using KDE Plasma5 on Slackware64-current every day, the whole day, and have no issues.
      My desktop (which I use more at the moment than my laptop) has a Nvidia GPU and I use their binary drivers. Also I added the PAM support from the Slackware (and ktown) testing area.
      Baloo can be resource-intensive during the first hours on a freshly installed system while it is busy indexing your files, but that will go back to normal once the indexing is complete. Sometimes a baloo index becomes corrupted (for instance when you do a forced power-down) and then you need to stop baloo, delete its index and restart baloo. But issues with baloo should normally not prevent you from logging in. Is your harddisk having physical issues perhaps?

      To remove and re-create the baloo database, try:
      # balooctl stop ; balooctl disable ; rm -rf ~/.local/share/baloo ; balooctl enable ; balooctl start

      1. Hi again.

        No, the hard disk is a new one and apparently has no errors.

        When starting Plasma from the console it get stuck with this messages:

        Registering “:1.35/StatusNotifierItem” to system tray
        Registering “:1.78/StatusNotifierItem” to system tray
        Registering “:1.80/StatusNotifierItem” to system tray
        qt.qpa.xcb: QXcbConnection: XCB error: 3 (BadWindow), sequence: 1302, resource id: 98566147, major code: 7 (ReparentWindow), minor code: 0
        qt.qpa.xcb: QXcbConnection: XCB error: 3 (BadWindow), sequence: 1304, resource id: 98566147, major code: 141 (Unknown), minor code: 1
        qt.qpa.xcb: QXcbConnection: XCB error: 3 (BadWindow), sequence: 1305, resource id: 98566147, major code: 6 (ChangeSaveSet), minor code: 0
        qt.qpa.xcb: QXcbConnection: XCB error: 3 (BadWindow), sequence: 1306, resource id: 98566147, major code: 25 (SendEvent), minor code: 0
        qt.qpa.xcb: QXcbConnection: XCB error: 3 (BadWindow), sequence: 1307, resource id: 98566147, major code: 12 (ConfigureWindow), minor code: 0
        qt.qpa.xcb: QXcbConnection: XCB error: 3 (BadWindow), sequence: 1309, resource id: 98566147, major code: 12 (ConfigureWindow), minor code: 0
        qt.qpa.xcb: QXcbConnection: XCB error: 3 (BadWindow), sequence: 1310, resource id: 98566147, major code: 8 (MapWindow), minor code: 0
        qt.qpa.xcb: QXcbConnection: XCB error: 3 (BadWindow), sequence: 1311, resource id: 98566147, major code: 61 (ClearArea), minor code: 0
        Service “:1.78” unregistered
        Container window visible, stack below
        kdeinit5: Got EXEC_NEW ‘/usr/lib64/qt5/plugins/kf5/kio/’ from launcher.
        kdeinit5: preparing to launch ‘/usr/lib64/qt5/plugins/kf5/kio/’

        Does it tell you anything?


          1. And today it worked fine! ๐Ÿ˜ฎ

            I have no idea of what’s going on. I just installed the last couple of upgrades for current, but apparently nothing relevant for this problem.

            I don’t know, if it happens again I will try to investigate further.


  2. will a version of DevHelp become available soon for slackware64-current ?

    Thanks for all your slackware efforts !!


  3. Hi again,

    Plasma started to fail again and apparently SDDM was crashing:

    Apr 11 10:58:51 liet kernel: [ 55.238700] sddm-greeter[1401]: segfault at 1501e36389 ip 00007fa561ecd2c8 sp 00007fff18a4d440 error 4 in[7fa561c77000+2d3000]
    Apr 11 11:05:44 liet kernel: [ 468.103765] sddm-greeter[1542]: segfault at 30 ip 00007f7970bccabd sp 00007ffdb99525e0 error 4 in[7f7970b0c000+2d3000]
    Apr 11 11:16:53 liet kernel: [ 36.322246] sddm-greeter[1383]: segfault at 1 ip 00007f9ab9d8cabd sp 00007ffe74067440 error 4 in[7f9ab9ccc000+2d3000]

    I removed all Plasma packages and installed them again. I even removed all user specific Plasma and KDE configuration (didn’t make any sense anyway since the error happened before any user info was loaded). No success, it was still failing 90% of the times (but once in a while it worked and I could not find a patter).

    Finally I made a new fresh slackware-current installation and installed again your Plasma set, and it is working apparently fine.

    My guess, I installed something that is conflicting with Plasma, but I have no idea of what was it. So I am installing again one by one all my packages and documented the order and all changes, so if it happens again I will be able to identify the conflicting software.


  4. Dear Alienbob,

    First, I very much appreciate your work on Slackware. I would like to suggest an improvement on Slackware installer, or at least ask a question for clarification.
    I just would like to ask question about the setup “Select Prompting Mode” step. The menu and the options have been changed, and I think the older version was better.
    In an older version of Slackware Install the options and the explanations were these:

    menu Choose groups of packages from interactive menus
    expert Choose individual packages from interactive menus.

    I’m referring to the old installer, please see on the Slackbook webpage 3.4.7:

    On the other hand, the newer and current version of Installer says this:

    menu Choose individual packages from interactive menus
    expert This is actually the same as the “menu” option

    I think the old version was better: it was more clear and it offered two genuinely different options: selection of groups of packages vs selections of individual packages. Why it has been changed?
    Also, I think it would be better if the users had this option still. Is it possible to, say, in next release of Slackware we’ll get the old intall options again with two different and clearly distinguishable options?

    In sum, I think the current installer is confusing or redundant. And I don’t see at all an improvement here from the newer version.

    If there is currently no real difference between the “menu” and “expert” options, then there is no reason to keep both of them. I would keep only one of those with the explanation: “Choose individual packages from interactive menus”.
    If, however, there is or should be a real difference and choice between individual vs groups of packages, then the old version should be used.

    I would go for the old version: the explanation is clear, it offers an extra option and more flexibility in installing.

    Thank you for clarification.

    Best wishes,

    1. Hi Peter,

      This blog is not the right place to make requests for a change in Slackware. We can discuss it here, no problem, but if you want to have attention from Pat Volkerding, be sure to post it in this LQ thread:

      This change to the setup was made before I created my git repository which has been tracking -current development since June 2018 ( for instance), so this was done a long time ago. But I seem to remember that the reason was that “expert” installs just took a very long time due to all the individual packages having to be (de-)selected and this adds nothing of value.

  5. Eric, I used your amazing script (no options) to copy the 30 April versions of the slackware64 live ISOs (current and current-xfce) onto a pair of thumb drives (a cheap PNY and a higher-end SanDisk Cruzer Glide 3.0), but neither my Lenovo T420s and X120 recognized any bootable software and bounced back to their respective boot drive selector screens. After the initial failures, I used “dd if=/dev/zero of=/dev/sdc bs=1M count=100” to wipe any boot records from the drives and used again, but again, the computers refused to boot.

    However, when I used dd to copy the ISOs onto the PNY drive, both computers booted both versions immediately. I’m using the current 3.7G ISO right now on my X120e to write this.

    Let me know if you need more information or want to try different ISO files.

    1. Well, “no options” is not correct, you would at least have to provide the ISO as input and the USB device as output parameter… but no other options are required to create a bootable USB stick.
      If the process of letting write the ISO to the USB stick had no errors, then I can not tell you why the USB stick would not boot your computers. I have a T460 on which my own USB Live stick boots just fine. The Live stick works on all (not a large amount) of the BIOS and UEFI computers I tested.

  6. Way back on March 21, 2020, I wanted to boot my circa 2013 Macbook Air using Liveslack XFCE on flash drive & ran into the old Broadcom_sta problem – no driver.

    The rest of the story: I downloaded live-current & installed it on a SSD connected via USB. I SlackBuilt broadcom_sta on it, installed wicd since I’m not familiar with NetworkManager, & now I’m off to the races. Way cool. Except for wireless master mode. Oh well…

    I configured it with my fave fvwm window manager + config which has suited me fine-aree for a b’zillion years. Starts up way faster than heavier window managers. At least I’m using my covid time for something I’ll use…:-)

    Cheers, Another Bob

    1. You know, all of the Live ISOs except the small XFCE variant, already have a “broadcom_sta” module in their “optional” module subdirectory.
      So if you boot a live medium (except XFCE) and add “load=broadcom_sta” to the boot commandline in the syslinux or grub menu page, you will get that Broadcom support out of the box.

  7. Hi Eric,
    after upgrading to freetype-2.10.2 (and creating/installing a compat32 package) yesterday on my -current64 multilib system i ran into problems with some wine apps. I finally found out that the new freetype version seems to be dependent on brotli. Created a brotli compat32 package was the solution.
    Maybe it’s possible to add that package to your /multilib/current/slackware64-compat32/ repo when you have time to refresh it.

  8. Hi Eric,

    It seems that synchronization of Slackware current repositories didn’t work last night on
    The Changelog entries dated “Tue May 26 20:35:03 UTC 2020” do not exist on the server at time of writing.

    1. Some of the cron jobs (not all apparently, just those for root) stopped working that day. I need to find out what happened, but I killed crond and restarted it, let’s see if that is sufficient.

  9. Here I go again – back on May 8 I described my success using live-current installed on a SATA SSD & booting it via USB cable – a SATA-USB2 cable. Works fine, lasts a long time!

    But I acquired an SATA-USB3 cable thinking it might be faster plus if my SATA-USB2 becomes disabled I would have an alternative. USB3 cable works fine when plugged into a running system – either Linux or MacOS. But it fails to boot.

    Boot starts successfully to the Live Slack screen & continues booting generic. Ultimately it stops with “SLACKWARELIVE: No live media found…trouble ahead” & suggests I add rootdelay=20 to boot command. It then drops to /bin/sh suggesting “RESCUE mode”. I tried various rootdelay’s with same result. It appears that the USB3 (I bought 2 & both fail to boot) never “sees” sdb & it’s partitions.

    I’ve been googlin’ around & looking at & haven’t figured anything out yet.

    The USB3 cable is described thus when plugged into running Linux:
    Bus 002 Device 004: ID 174c:55aa ASMedia Technology Inc. Name: ASM1051E SATA 6Gb/s bridge, ASM1053E SATA 6Gb/s bridge, ASM1153 SATA 3Gb/s bridge, ASM1153E SATA 6Gb/s bridge

    I booted liveslack using SATA-USB2 – success. I then plugged an unused SSD using SATA-USB3 – success. The USB3 cable I bought won’t support boot for some reason.

    Any suggestions?

    Much Obliged, Bob

    1. You’ll have to revert to troubleshooting in that rescue shell.
      Try stuff flike “fdisk -l” in that rescue shell, to see what the kernel detected and whether your USB drive is actually there, because if it is there then liveslak has a problem itself.
      You could also add “debug=4” to the boot commandline of liveslak. That will show a lot more output on your screen than during a regular boot and it will eventually stop in the emergency shell as well. But perhaps the added verbosity brings some clarity into what’s happening.

  10. I apologize for not providing more information.

    The RESCUE mode occurs after usbcore has registered usb-storage & usbhid modules. USB system has identified a Logitech mouse unifying receiver & an internal webcam as noted in “dmesg”. Not captured in “dmesg” are usb related errors which I will try to summerize below. The SSD partitions are not successfully discovered nor listed in /dev/disk/*. As noted by Ricardo above, lsmod reveals modules xhci_pci & xhci_hcd are both loaded, as are sdhci_pci, cqhci, sdhci, mmc_block, & mmc_core among others. Note my 2 linux laptops (kernel 5.4.4X) have only USB2 hardware while my Macbook Air has USB3; boot problem is essentially identical. “debug=4” doesn’t seem to reveal more information about the problem.

    Not during each boot, but occasionally, repeating messages such as these roll up the screen continuously. The device number XX increments by 1 in each group of errors. I surmise the USB system at boot time apparently isn’t happy with recognizing the USB3 cable.

    usb 1-1-port 1: unable to enumerate USB device
    usb 1-1.1: new full-speed USB device number XX using ehci-pci
    usb 1-1.1: device not accepting address XX. error -32
    usb 1-1-port 1: attempt to power cycle
    usb 1-1.1: new full-speed USB device number XX using ehci-pci
    usb 1-1.1: device descriptor read/8. error -32
    usb 1-1.1: device descriptor read/8. error -32

    I can’t seem to interrupt the scrolling thus what I have typed is questionable given my old eyes trying to focus on it while flying up the screen. I think usb 1-1.1 is a hub & not the USB3 cable which I believe is usb 2-1.2.

    Another tidbit: when successfully booting using the USB2 cable, I see this:
    [ 19.839615] EXT4-fs (sdb3): mounted filesystem with ordered data mode. Opts: (null)
    [ 20.886272] /dev/loopX: Can’t open blockdev
    …[above line repeats for X=0..19]]
    [ 21.818404] EXT4-fs (sdb3): re-mounted. Opts: (null)

    All is curious, don’t know if it’s worth pursuing at length. Perhaps a kernel 5.4.4X vs liveslak’s kernel 5.4.35?

  11. Upgraded to with kernel 5.4.42 – same problem. I’ve ordered a different SATA-USB3 cable. Will post report.

  12. Assuming this will get seen and is the right place for this!

    I tried out slackware live. The ISO I downloaded worked OK except I noticed an error scrolling by at boot time. It is a “chown” command that uses the –reference argument in liveinit.tpl. Since the boot process uses “busybox” there is no –reference argument. That didn’t seem to create problems, but it is there.

    Unfortunately when I tried to make my own ISOs, even with just default settings, I would be unable to log in with either account. I would get strange “incorrect” errors on the GUI login page, and when I tried the console it gave “permission denied” errors (for both accounts). I could not figure out what the issue was.

    I don’t think it was the chown issue above, because I did modify liveinit.tpl. So I’m at a loss.

    This is not a crucial thing for me. I was just experimenting with the possibility of making custom ISOs for live USB sticks, which would be very cool. (The project as a whole seems very cool except for the above problem.)

    1. Hi Fred,

      Can you share the exact command you used to create an ISO? Perhaps I can give you some pointers. The issue sounds like PAM-inflicted, so perhaps you pointed the ‘’ script to a slackware-current mirror that was in the process of updating itself?
      And thanks for pointing out the busybox related error in liveinit.tpl. I will have a look into that.

  13. Hello,

    Managed to get the process to work by removing and updating everything. I have a local mirror that I update with the script you wrote, and probably my mirror was not consistent or something.

    Anyway thanks, everything seems to work fine now!

    1. Good to hear Fred.
      If you run into issues during your Live ISO customization, let me know so that I can see whether it is fixable through liveslak modifications.

  14. Re: Booting from SSH using SATA to USB3 cable failure.

    I bought a Sabrent SATA-USB3 cable & no change in boot failure behavior previously described. rootdelay=60 causes a notable pause but boot still fails. Yet booting with a SATA-USB2 cable succeds.

    Appreciate all you do. Next I will try a USB3.0 & USB3.1 flash drive i.e. not an SSD connected via SATA-USB3 cabke,

    1. I tried liveslak on USB3.0 & USB3.1 flash drives. They boot properly & indeed they are quicker than my USB2 flash drives – expected that of course. I’ve had flash drives “die” unexpectedly which is why I was working with a SSD connected via SATA-USB3 cable (see above). I’ll update liveslak when available on my SSD & continue trying to boot using the USB3 cable. Will post results.

      Cheers to All, Bob

  15. Hi Eric! I just got your LibreOffice 6.4.5 packages and all work OK. However it seems like the kde-integration packages are missing. Are they no longer included? Thanks!

  16. Hi Eric! Saw that the Plasma5 support for LibreOffice 6.4.5 was added today. Thank you!!

    I can confirm it works OK. Stay safe and get well,


  17. Hi Erric

    thanks for all the job you do. I have two questions, is there any particular reason you kept calibre2 in you repo? the same also on vlc, vlc2 has any reason?

    The reason I am aksing is that the names vlc2 or calibre2 sound like there are now new vlc or calibre and these are the new product name. Also slackpkg info (.txt or .meta) does not give any clue about the version, or why one shall decide for one or the other.

    A friend of mine ask me these & I ended up opening the repo and checking the date of package, is there any other way you suggest?


    1. In my repository, ‘vlc’ and ‘calibre’ will give you the preferred version of the package that you should install for your version of Slackware Linux.
      Sometimes I make alternative releases available because people prefer that particular version instead of the recommented version.
      So, calibre2 and vlc2 offer an older release of calibre and vlc if you would want that.

  18. Just wanted to say THANK YOU!
    Your VLC build script is a marvel, it has indirectly helped power SiriusXM radio transmissions for a number of years.

    Thanks Alien Bob, hope your ship is repaired soon enough to fly back out and travel the Universe.

    Your talent is much appreciated Sir

  19. Just wondering, I know you are quite busy with other matters. Who has picked up kde/plasma? They’re up to Frameworks 5.72.0 and Plasma 5.19.3 with no Slackware July update. How do you get your source packages for deps and kde in your KTown repository? I think your SlackBuild scripts would work if I could get the new sources.

    1. You could also just enjoy your current installation. If new packages get released you’ll find out soon enough.
      If you do not know how to get the source tarballs for newer releases, then it’s best to wait in any case.

  20. I installed Kde plasma in Slack current and the shutdown options do not appear in the menu, I had plasma installed before and it worked well, but since it was updated it does not appear shutdownfile:///home/portu/Imรกgenes/Screenshot_20200804_150918.png

    1. My guess is that you forgot “slackpkg install-new” and missed the new mozjs68 package which is required to make polkit work. And a non-working polkit causes exactly the issues you encounter.

      You could easily have found out yourself what was wrong, if only you would have read my most recent blog post about KDE Plasma5 and especially the comments below that article. The article is only a few days old and hard to miss – it covers the blog’s front page.
      This generic feedback page is not the right place to report this kind of PLasma5-specific issues. Also I do not know what you wanted to achieve with that screenshot? It is not a working link.

  21. Seems like there might be an issue with liveslak in CINNAMON variant.
    Making a CINNAMON ISO produces a system which cannot be logged in.
    Credentials both for ‘root’ and ‘live’ accounts are accepted but then ‘Permission denied’ is immediately given, when either in X display manager screen or in a console.
    (ISOs in both SLACKWARE and PLASMA5 variants are okay.)

          1. Capital!
            Thank you, Eric, and thank you, Willy!
            I’ll try it ASAP (hoping there’ll be nothing negative to report).

          2. @Eric
            I suppose you’d want to correct the line 99 in
            Now it reads (with doubled dblquote):
            LIVEUIDFN_DAW=””${DISTRO^} Live Musician”
            I don’t know advanced bash, but the script won’t start with that.

  22. The Changelog for Sept. 15, 2020 18:51 shows:
    “l/mozjs78-78.2.0esr-x86_64-1.txz: Added.
    This is needed by the new version of polkit.
    l/polkit-0.118-x86_64-1.txz: Upgraded.
    This requires the new mozjs78 package.”

    However the upgrade via slackpkg did not include polkit-0.118.
    So the package was specifically upgraded to 0.118.
    Today when the system was upgraded the polkit packages was rolled back to 0.117.

    I must have missed a notification. Any ideas/clarification?

    1. Andrew, this is normal when you use slackpkg+ to manage 3rd-partyt repositories: my ‘ktown’ version of polkit has precedence over the Slackware polkit package, and so you’ll always end up with my version 0.117 after a “slackpkg upgrade-all”.
      The Slackware version of that new polkit package does not include elogind support so you’ll have to stick with the older ktown version (and keep mozjs68 installed!) or else recompile polkit-0.118 using my modified SlackBuild script to enable elogind support again.

      1. I’ll have to go with the latter.
        Slackbuild, 120: –enable-libelogind=yes
        dep: elogind-243.7-x86_64-1alien
        be sufficient?

  23. Hello Eric,

    In the latest Slackware64 (kernel 5.4.66) KDE 5.20.08 I faced with next problems:
    Can’t connect to my WiFi network from KDE Networks. After clicking on connect button, “Not authorized to control networking.” message appears. But my user is added to netdev group.
    Also K Menu starts freezing right after system booting and after a while works good.

    Best regards,

    1. I have articles on this blog specifically discussing KDE Plasma package releases. The “your feedback” page really is not the place for questions like this, since you may miss the information that could have helped you answer the question all by yourself.
      Did you read the Slackware ChangeLog.txt recently? Noticed the polkit update? Are you aware what consequences this has for a Plasma5 installation if you blindly update Slackware?

  24. Thank you for making “slackware current” available. I love it and can use it on my laptop. Thank you and continue this great job. Kisses.

  25. Hi Eric

    Can you please kindly give me hint if I do a mistake here?

    Because of my printer, I am using glibc-solibs from your multilib repo. I am on the current version. Lately, when I installed the “glibc-solibs-2.30_multilib-x86_64-1alien” to get the 32-bit driver of my printer work, I ended up in a terrible situation, shell was gone, and I could not even boot, if I were not wrong /bin/sh was not working anymore. Even chroot worked only when I bound the lib64 folder of the live image. So, I ended up using a live image and correct bunch of lib64 libraries and revert the links. Any clue what I am doing wrong?

    I even found last week something similar here
    so I am not the only one who comes up with the Idea to take only glibc-solibs. Do you have any clue? Is something changed?

    Thanks in advance

    1. Are you sure you have a 64bit Slackware installation? If you installed the multilib glibc-solibs, and then stuff broke so bad you were unable to boot, and you needed a live ISO to fix a lot of lib64/ libraries, then the most likely explanation is that you have been mixing 32bit and 64bit Slackware packages, or even replacing 64bit packages with their 32bit versions..
      That is never the correct way to fix problems. A multilib add-on to a 64bit Slackware leaves ALL the Slackware 64bit packages un-touched. All it does is ADD a bunch of converted 32bit packages. With the exception of glibc (and gcc if you also want to be able to compile 32bit packages).
      If you are actually running Slackware 64bit, then it may enlighten you if you boot a Live OS and do something like “ls /var/log/packages/ |grep -E ‘i.68′” to find out how many 32bit packages are actually installed on your 64bit OS. The number should be ZERO.

      1. I am using a pure x86-64 current, at least it is what I told myself the last 10 years. Let us hope I am not that stupid. ๐Ÿ˜€ Here is also the output for your command:
        20:38:52 >> $ ls /var/log/packages/ |grep -E โ€˜i.68โ€ฒ | wc -l
        My observation is the list below will every time get replaced in /lib64 with the one from the package. Effectively these library files will be moved to the /lib64/incoming folder and the new ones will be placed in /lib64, linked to, and get used. Is it correct behavior and what you will expect? Excuse my ignorance isn’t it suppose to be placed in /lib and then linked from there? or the /lib64 is really the place for multi-lib files?

        The lib64 is so wrong that I need to bind /lib64 of host for chrooting, else it complains, that it cannot find /bin/bash even, while /bin/bash is obviously there. My call is some basic commands are not working anymore! While booting, it complains even earlier and about /bin/sh.

        Obviously when I correct /lib64, linking back to Patrick’s lib64 files, chrooting is fine and I can do simple chroot, with only proc, system and device bound, and also the installation boot. Any other guess, am I using the correct package at all?

        Side note: I solved my problem with the printer driver hacking and creating a pure 64 bit base on another driver, but I would still appreciate your advice on what could be my mistake. I may need this for other stuff, and I would like to avoid going full multi-lib, jus to keep things simple for myself.

        Thanks in advance.

        1. Inman, I think you do not realize what “multilib” means.
          In a pure64 bit system you will have only 64bit binaries, and libraries placed in /lib64 and /lib64. On a pure 32bit system you will only have 32bit binaries and libraries placed in /lib and /usr/lib.
          A multilib system is in its core a 64bit system with the added capability of (compiling and) running 32bit binaries. Some original core OS packages like glibc-* (for runtime support) and gcc-* (for compile-time support of 32bit) are going to be replaced with multilib versions. That means, these multilib versions of glibc and gcc packages contain both 64bit and also 32bit binary code, and you will find that their lbraries will be placed in all of /lib64 , /lib , /usr/lib64 and /usr/lib.
          Then you’ll need runtime support for 32bit programs that you want to (compile or) run. Every program links to shared libraries at runtime, and for a 32bit program you’ll have to provide the 32bit runtime shared libraries. This part is fulfilled on Slackware64 multilib by my set of “compat32” packages which are nothing more than original Slackware32 packages, re-packaged with a slighly different package name (adding “-compat32” to the name and using x86_64 architecture even though the binaries are 32bit).

          It looks like you (correctly) replaced Slackware’s glibc-solibs with my multilib version. However, if you have other Slackware glibc packages installed, you need to replace them all, not just -solibs!

          If you replace the glibc packages with their multilib versions, a chroot should not stop working. If you have errors like “cannot find /bin/bash” that actually means that a bash binary is found but it is the wrong architecture. Which leads me to believe (also considering all that symlinking you thought was needed) that you removed the 64bit glibc libraries or at least made the 32bit libraries found first in the library path.

          That will only break stuff and is completely unneeded.

  26. Hi Eric,
    I noticed you have an opencv package in the ktown deps, what is it needed for?
    I usually build my own from SBo but i see you use my same options so i don’t mind borrowing it now that i use a weak laptop. Do you build it with OpenBLAS support? also do you plan to keep building it now that ktown is not getting updates?

    1. OPencv is required by Digikam and can be used with KDEnlive as well. I do not compile it against OpenBLAS, which I only have in my regular repository, not in ktown. I have to tread carefully when it comes to dependencies that may prevent software inclusion in Slackware at some future date.
      I am no longer providing updates to ‘ktown’ at this moment (except security updates and fixing broken packages).
      I need Patrick Volkerding to adopt Plasma5 into Slackware and when he does that, I will see what packages he did not accept. Then I will decide how to proceed with the abandoned stuff.

  27. Hi AlienBob,

    I would like to say thank you very much for your blog-posts, for often really helpful informations and the packages especially for the 32-bit systems. All of that has saved my time more then once.


    1. Mat, you’re welcome. And good to know that there are still people who use 32bit Slackware. I sometimes wonder, is it still worth the effort creating these 32bit packages, until I get a message like yours here.

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.