#archlinux32 | Logs for 2019-10-12

Back
[00:05:17] <buildmaster> i686/llvm is broken (says nlopc46): https://archlinux32.org
[00:20:45] <bill-auger> i notice while upgrading my arch32 i686 system that it is pulling in mostly .pentium4 packages - is that expected
[01:11:41] <bill-auger> also makepkg fails with "not available for the pentium4 architecture"
[01:12:14] <bill-auger> i could probably change that to pentium4; but i am trying to build for the i686 architecture
[01:12:39] <bill-auger> maybe i messed up my pacman.conf ?
[01:30:55] <djmoch> there was a pacman update a while back that changed the way Architecture = auto behaves. for systems that support pentium4, it switched automatically.
[01:31:48] <djmoch> bill-auger: ^
[01:31:57] <djmoch> i got caught with the same thing a while back
[01:33:02] <djmoch> looks like it switched back in May, if you check the homepage
[01:34:09] <bill-auger> i see - yea i have not upgraded it in a while
[01:34:27] <bill-auger> theres probably now way to rollback to i686?
[01:35:05] <bill-auger> i could re-install again from a i686 ISO but how would i lock it to i686
[01:35:36] <djmoch> set Architecture = i686 in pacman.conf
[01:36:02] <bill-auger> awesome thanks
[01:36:04] <djmoch> then there's some pacman magic you'll have to do to make sure you've got the right packages installed
[01:36:19] <djmoch> maybe just pacman -Syyu
[01:37:03] <djmoch> a bit of advice, though, unless you have a reason for wanting to use i686, you should probably just go with pentium4
[01:37:11] <djmoch> especially if you're running X
[01:37:16] <bill-auger> ok i see pacman.conf had Architecture = Auto
[01:37:35] <bill-auger> yes i have specfic reason
[01:37:53] <djmoch> okay then. carry on. :)
[01:38:09] <bill-auger> i use it to debug packages when something in parabola breaks
[01:39:13] <djmoch> interesting. i know parabola and arch have a lot in common, but why would you use the latter to debug an issue with the former?
[01:40:02] <bill-auger> to determine of the bug is only present in parabola or if it is also present in arch32 - that determines where to post the BR
[01:40:31] <djmoch> i see
[01:40:44] <bill-auger> today i have a problem building the calibre PKGBUILD - the python setup.py command segfaults
[01:42:17] <djmoch> yeah, i have to think a segfault would be a problem with the interpreter and not calibre per se
[01:42:44] <djmoch> but still useful to know if it works in arch, but not parabola
[01:42:56] <bill-auger> right but parabola's python comes from arch32
[01:43:21] <bill-auger> its not co-incidentally "a lot in common" - parabola uses arch32 as its upstream, and replaces a few of them, such as calibre
[01:43:38] <djmoch> i didn't know that
[01:43:40] <bill-auger> *replaces a few packages
[01:51:20] <bill-auger> i dont think i can migrate back to i686 - pacman -Syyuu only replaced 12 packages - but i just upgraded over 300 packages , almost all were pentium4
[01:52:37] <djmoch> hmm ... i'm not sure
[01:52:51] <bill-auger> and makepkg still wont start
[01:53:14] <bill-auger> uname -a says it is running i686 kernel now - so i dont suppose rebooting would help
[01:53:38] <djmoch> try pacman -Qqn | pacman -S
[01:54:13] <djmoch> i just realized it probably won't try to upgrade if the same version is in both pentium4 and i686
[01:54:46] <djmoch> sorry pacman -Qqn | pacman -S -
[01:54:52] <djmoch> forgot the trailing -
[01:57:09] <bill-auger> /etc/makepkg.conf had pentium4 hard-coded in a few places
[01:57:41] <buildmaster> pentium4/llvm is broken (says buildknecht2): https://archlinux32.org
[01:58:57] <djmoch> bill-auger: i don't think that should affect the behavior of pacman
[01:59:30] <bill-auger> djmoch: thats was a good idea - it is replacing 500+packages now
[02:00:34] <bill-auger> i did not affect pacman - i am trying to run makepkg - makepkg still failed with "not available for the pentium4 architecture"
[02:01:02] <bill-auger> i expect everything will be fine now (or totally broken)
[02:26:26] <bill-auger> yay its all good - thans djmoch
[03:03:16] <bill-auger> eschwartz: same segfault on arch32 - both with calibre v3.48.0 and also 4.1.0 https://bugs.archlinux32.org
[03:03:18] <phrik> Title: FS#91 : [python2]: segfault (at bugs.archlinux32.org)
[03:10:28] <bill-auger> unfortunately i can not see if arch32 has any special patches for calibre - cause oops 500 https://www.archlinux32.org
[03:12:06] <bill-auger> co-incidentally a python2 error ^
[03:13:28] -!- yans has joined #archlinux32
[03:14:28] <bill-auger> oh actually the patches are in a gitlab instance IIRC maybe i can find that
[03:16:54] <bill-auger> nope the only patch is to the check() function https://git.archlinux32.org
[03:16:56] <phrik> Title: archlinux32/packages: Package customizations and pure-i686 packages - Archlinux32 Gitea (at git.archlinux32.org)
[03:33:38] -!- thePiGrepper has quit [Ping timeout: 245 seconds]
[03:45:51] -!- thePiGrepper has joined #archlinux32
[04:12:13] -!- BeepBeepJeep has joined #archlinux32
[04:12:28] <BeepBeepJeep> Hello, Trying to bootstrap from debian to do an install
[04:12:38] <BeepBeepJeep> This script has a few errors in it https://github.com
[04:12:40] <phrik> Title: arch-bootstrap/arch-bootstrap.sh at i686-support · tokland/arch-bootstrap · GitHub (at github.com)
[04:12:48] <BeepBeepJeep> ive been able to fix the one... with an extra ;
[04:13:05] <BeepBeepJeep> but its not going through the list of packages like it should
[04:13:08] <BeepBeepJeep> plz hlp
[04:14:28] <BeepBeepJeep> may have been a user issue lol
[04:15:08] <BeepBeepJeep> wasnt running as root
[04:15:15] <BeepBeepJeep> didnt have correct permissions
[04:15:18] <BeepBeepJeep> for tmp
[04:19:12] <BeepBeepJeep> seems like theres a bunch of issues
[04:19:16] <BeepBeepJeep> https://bbs.archlinux32.org
[04:19:18] <phrik> Title: arch-boostrap.sh error i686 / Installation / Arch Linux 32 Forums (at bbs.archlinux32.org)
[04:19:52] -!- samantaz__ has quit [Ping timeout: 264 seconds]
[04:34:47] <BeepBeepJeep> chroot_args=()
[04:34:47] <BeepBeepJeep> [[ $userspec ]] && chroot_args+=(--userspec "$userspec")
[04:34:47] <BeepBeepJeep> SHELL=/bin/bash unshare --fork --pid chroot "${chroot_args[@]}" -- "$chrootdir" "$@"
[04:35:10] <BeepBeepJeep> ./arch-chroot /mnt
[04:35:11] <BeepBeepJeep> unshare: unshare failed: Invalid argument
[04:39:57] -!- yans has quit [Ping timeout: 240 seconds]
[04:49:05] -!- BeepBeepJeep has quit [Remote host closed the connection]
[05:21:32] <buildmaster> i686/hedgewars are broken (says rechenknecht) - already flagged out-of day upstream on 2019-10-09: https://archlinux32.org
[05:23:16] <buildmaster> pentium4/hedgewars are broken (says buildknecht2) - already flagged out-of day upstream on 2019-10-09: https://archlinux32.org
[07:00:51] -!- thePiGrepper has quit [Ping timeout: 240 seconds]
[07:06:31] -!- thePiGrepper has joined #archlinux32
[08:24:53] <nit-picker> package projectm-pulseaudio-3.1.0-5.5-pentium4.pkg.tar.xz has an invalid signature.
[09:02:37] -!- deep42thought has joined #archlinux32
[09:02:37] <buildmaster> Hi deep42thought!
[09:02:37] <buildmaster> !rq deep42thought
[09:02:39] <phrik> buildmaster: <deep42thought> my supervisor when I made my first shot with the laser: He handed me the remote and said: "Hey Erich, want to destroy a 20 Mio€ device?"
[09:03:02] <deep42thought> City-busz_: I removed libmagick and rescheduled converseen
[09:03:18] <deep42thought> I'll take a look at the scheduling of any packages - your suggestion makes sense.
[09:03:53] <deep42thought> bill-auger: if "pentium4" appears in makepkg, it indicates, that you have installed pacman for pentium4
[09:04:06] <deep42thought> have you merged your new config files after reinstalling for i686?
[09:04:22] <deep42thought> s@makepkg@/etc/makepkg.conf@
[09:04:58] <bill-auger> deep42thought: i did - we figured it out
[09:05:08] <deep42thought> ah, ok, then: good :-)
[09:05:29] <bill-auger> my pacman.conf had ARchitecture = Auto - and -Syu converted my system to pentuim4
[09:05:36] <deep42thought> yes
[09:05:44] <deep42thought> that part was clear to me from the log
[09:06:00] <deep42thought> the question is: does your makepkg happily build i686 packages, now?
[09:06:08] <bill-auger> i rolled it back successfullly with pacman -Qqn | pacman -S -
[09:06:48] <bill-auger> i edited makepkg.conf by hand - after that, makepkg started
[09:08:13] <bill-auger> it had pentium4 hard-coded in a few places - i thnk, because i installed it (base-devel) after i was (unknowingly) running pentium4 system
[09:12:39] <deep42thought> `pacman -Qqn | pacman -S -` should have rolled back the makepkg.conf, too
[09:12:50] <deep42thought> you may need to merge the pacnew file, though
[09:12:57] <bill-auger> it may have
[09:13:39] <bill-auger> in the end thats all fine, the reason i was doing it was to show that python2 is broken
[09:14:17] <bill-auger> oh i actually i did check the REPL - that worked
[09:14:23] <bill-auger> something is wrong with it though
[09:29:58] -!- infides has joined #archlinux32
[09:37:35] -!- infides has quit [Ping timeout: 250 seconds]
[10:15:40] -!- thePiGrepper has quit [Ping timeout: 265 seconds]
[10:20:40] <City-busz_> deep42thought: thanks. I think you can schedule nearly all 'any' packages on the broken builds list, becasue most of them are failed due to missing dependencies in a i486 chroot.
[10:21:05] <deep42thought> I did not yet change the any scheduling logic
[10:21:25] <deep42thought> because it's currently just the generic case "build wherever you can, prefer identical architecture"
[10:21:44] <deep42thought> I have to think a little longer, how to fit the "any" case into that
[10:22:24] <City-busz_> All of these packages: https://www.archlinux32.org
[10:22:27] <phrik> Title: Arch Linux 32 - List of Package Builds (at www.archlinux32.org)
[10:22:51] <City-busz_> deep42thought: or you have to blacklist all of these packages for i486.
[10:23:58] <deep42thought> no, that's not possible
[10:24:15] <deep42thought> or, well, it *is* possible, but it will not prevent building the packages on i486
[10:25:16] <City-busz_> then you have to build them in pentium4. :) Or try in i486, then try in i686, and finally in pentium4 before mark them as broken package
[10:27:03] <City-busz_> but I think it would be simplier to build them on pentium4, where all dependencies are available.
[10:27:30] <deep42thought> yes
[10:27:45] <deep42thought> as I said: it makes total sense to build them on pentium4 *only*
[10:27:53] <City-busz_> okay :)
[10:27:59] <deep42thought> but I have to think, how to best accomplish that (without breaking other stuff)
[10:30:00] <deep42thought> ok, I said the buildmaster, that "'any' and 'i486' are not "build-slave-compatible", e.g. one cannot be built on the other - hopefully, this is only queried in get-assignment and nothing else breaks :-)
[10:32:07] <City-busz_> I think you can blacklist packages that depend on 'electron': caprine, min, react-native-debugger, riot-desktop, wire-desktop, cozy-desktop
[10:32:30] <deep42thought> right
[10:34:47] <City-busz_> deep42thought: why do I get error, when I visiting this page? https://www.archlinux32.org
[10:35:12] <deep42thought> because the mysql database is inconsistent with the pacman database
[10:35:13] <deep42thought> :-(
[10:41:59] <buildmaster> i686/ldc is broken (says eurobuild6-6): https://archlinux32.org
[11:14:57] -!- davor has quit [Ping timeout: 240 seconds]
[11:16:23] -!- davor has joined #archlinux32
[11:17:29] <buildmaster> i686/zig is broken (says eurobuild6-4): https://archlinux32.org
[11:19:25] <buildmaster> i686/intel-opencl-clang is broken (says rechenknecht): https://archlinux32.org
[11:21:25] -!- buildmaster has quit [Remote host closed the connection]
[11:21:37] -!- buildmaster has joined #archlinux32
[11:21:38] <buildmaster> !rq buildmaster
[11:21:39] <phrik> buildmaster: <buildmaster> I might be insane, but never confused ... ;-)
[12:50:41] <City-busz_> deep42thought: on pentium4, please move cheese, gnome-contacts and totem from [testing] to [extra] to fix broken linking on libgnome-desktop.so
[13:01:20] <deep42thought> City-busz_: moving now
[13:01:49] -!- djmoch has quit [Quit: ZNC 1.7.4 - https://znc.in]
[13:02:13] <deep42thought> done
[13:02:41] -!- thePiGrepper has joined #archlinux32
[13:06:17] -!- djmoch has joined #archlinux32
[13:07:07] -!- deep42thought has quit [Quit: Leaving.]
[13:12:06] <buildmaster> i686/qtcreator is broken (says eurobuild6-1): https://archlinux32.org
[13:21:45] -!- samantaz__ has joined #archlinux32
[13:51:20] <buildmaster> i486/facile is broken (says eurobuild6-7-i486): https://archlinux32.org
[14:02:03] <buildmaster> i686/facile is broken (says eurobuild6-2): https://archlinux32.org
[14:02:39] <buildmaster> pentium4/facile is broken (says eurobuild6-4): https://archlinux32.org
[14:13:35] <buildmaster> i686/intel-graphics-compiler is broken (says eurobuild6-6): https://archlinux32.org
[14:36:51] -!- thePiGrepper has quit [Ping timeout: 240 seconds]
[14:37:45] -!- thePiGrepper has joined #archlinux32
[16:26:05] -!- Scimmia has joined #archlinux32
[16:33:18] -!- Scimmia has parted #archlinux32
[17:36:45] -!- niki has quit [Remote host closed the connection]
[17:40:28] -!- niki has joined #archlinux32
[17:51:21] -!- thePiGrepper has quit [Quit: leaving]
[18:43:19] <buildmaster> i486/nodejs-lts-carbon is broken (says eurobuild6-7-i486): https://archlinux32.org
[19:05:32] -!- oaken-source has joined #archlinux32
[22:57:31] <buildmaster> any/man-pages are broken (says eurobuild6-2): https://archlinux32.org
[23:14:59] -!- samantaz_ has joined #archlinux32
[23:18:06] -!- samantaz__ has quit [Ping timeout: 240 seconds]