#archlinux32 | Logs for 2020-09-05

Back
[00:24:13] -!- charims_ has joined #archlinux32
[00:28:56] -!- charims_ has quit [Client Quit]
[00:30:06] -!- charims_ has joined #archlinux32
[00:32:17] <buildmaster> pentium4/haskell-binary-orphans are broken (says nlopc46): https://archlinux32.org
[00:35:14] <buildmaster> i686/haskell-binary-orphans are broken (says rechenknecht): https://archlinux32.org
[00:36:17] <buildmaster> i686/haskell-bytestring-tree-builder is broken (says nlopc46): https://archlinux32.org
[00:40:06] <buildmaster> i686/haskell-cborg is broken (says nlopc46): https://archlinux32.org
[00:42:52] <buildmaster> pentium4/haskell-bytestring-tree-builder is broken (says rechenknecht): https://archlinux32.org
[00:43:54] <buildmaster> pentium4/haskell-cborg is broken (says nlopc46): https://archlinux32.org
[00:47:30] <buildmaster> pentium4/haskell-edit-distance-vector is broken (says nlopc46): https://archlinux32.org
[00:49:54] <buildmaster> i686/haskell-edit-distance-vector is broken (says rechenknecht): https://archlinux32.org
[00:51:10] <buildmaster> i686/haskell-extra is broken (says nlopc46): https://archlinux32.org
[00:57:47] <buildmaster> pentium4/haskell-extra is broken (says rechenknecht): https://archlinux32.org
[00:58:43] <buildmaster> i686/haskell-http-types are broken (says nlopc46): https://archlinux32.org
[01:02:24] <buildmaster> i686/haskell-interpolate is broken (says nlopc46): https://archlinux32.org
[01:04:10] <buildmaster> pentium4/haskell-http-types are broken (says rechenknecht): https://archlinux32.org
[01:06:12] <buildmaster> pentium4/haskell-interpolate is broken (says nlopc46): https://archlinux32.org
[01:09:53] <buildmaster> pentium4/haskell-parsers are broken (says nlopc46): https://archlinux32.org
[01:11:16] <buildmaster> i686/haskell-parsers are broken (says rechenknecht): https://archlinux32.org
[01:13:37] <buildmaster> i686/haskell-prettyprinter is broken (says nlopc46): https://archlinux32.org
[01:17:13] <buildmaster> i686/haskell-text-short is broken (says nlopc46): https://archlinux32.org
[01:17:45] <buildmaster> pentium4/haskell-prettyprinter is broken (says rechenknecht): https://archlinux32.org
[01:20:52] <buildmaster> pentium4/haskell-text-short is broken (says nlopc46): https://archlinux32.org
[01:54:13] <buildmaster> i686/haskell-simple-sendfile is broken (says nlopc46): https://archlinux32.org
[01:56:44] <buildmaster> i686/haskell-mono-traversable is broken (says rechenknecht): https://archlinux32.org
[01:57:49] <buildmaster> pentium4/haskell-simple-sendfile is broken (says nlopc46): https://archlinux32.org
[02:37:26] <buildmaster> i686/haskell-safecopy is broken (says nlopc46): https://archlinux32.org
[02:41:08] -!- yans has quit [Ping timeout: 256 seconds]
[02:41:51] <buildmaster> pentium4/haskell-safecopy is broken (says rechenknecht): https://archlinux32.org
[02:44:27] <buildmaster> i686/haskell-haddock-library is broken (says nlopc46): https://archlinux32.org
[02:48:43] <buildmaster> pentium4/haskell-haddock-library is broken (says rechenknecht): https://archlinux32.org
[02:57:03] <buildmaster> i686/onednn is broken (says nlopc46): https://archlinux32.org
[03:01:50] <buildmaster> pentium4/onednn is broken (says rechenknecht): https://archlinux32.org
[03:16:55] -!- Alina-malina_ has joined #archlinux32
[03:17:44] -!- Alina-malina has quit [Ping timeout: 265 seconds]
[03:29:52] <buildmaster> i486/onednn is broken (says nlopc46-i486bs0): https://archlinux32.org
[03:34:56] <buildmaster> i486/lxd is broken (says nlopc46-i486bs0): https://archlinux32.org
[03:39:34] <buildmaster> i486/linux is broken (says nlopc46-i486bs0): https://archlinux32.org
[03:45:13] <buildmaster> i686/haskell-hslua is broken (says nlopc46): https://archlinux32.org
[03:49:23] <buildmaster> pentium4/haskell-hslua is broken (says rechenknecht): https://archlinux32.org
[04:15:35] <buildmaster> i686/haskell-binary-parser is broken (says nlopc46): https://archlinux32.org
[04:19:31] <buildmaster> i686/haskell-bytestring-strict-builder is broken (says nlopc46): https://archlinux32.org
[04:19:35] <buildmaster> pentium4/haskell-binary-parser is broken (says rechenknecht): https://archlinux32.org
[04:24:37] <buildmaster> i686/haskell-neat-interpolation is broken (says nlopc46): https://archlinux32.org
[04:26:22] <buildmaster> pentium4/haskell-bytestring-strict-builder is broken (says rechenknecht): https://archlinux32.org
[04:28:34] <buildmaster> pentium4/haskell-neat-interpolation is broken (says nlopc46): https://archlinux32.org
[05:33:42] <sunshavi> for archlinux-arm there is this repo for the pkgbuilds https://github.com
[05:33:43] <phrik> Title: GitHub - archlinuxarm/PKGBUILDs: PKGBUILDs modified to build on Arch Linux ARM (at github.com)
[05:33:54] <sunshavi> where is the equivalent for arch32?
[05:38:18] <bill-auger> for it to be equivalent, does it need to be on github?
[05:38:58] <sunshavi> bill-auger: not really. I just want to find it. for recompiling manually a package
[05:39:57] <bill-auger> the PKGBULD repo is self-hosted
[05:40:32] <bill-auger> its a bit funky though - you can not use them with makepkg - they are a set of patches on top of the arch ABS repo
[05:41:03] <bill-auger> if you want regular PKGBUILDs, it would be easier to use parabola's
[05:41:45] <bill-auger> https://www.parabola.nu
[05:41:46] <phrik> Title: Parabola GNU/Linux-libre - Package Search (at www.parabola.nu)
[05:42:05] <sunshavi> phrik: is a bot?
[05:42:05] <phrik> Negatively true...
[05:42:51] <bill-auger> parabola PKGBUILDs are a combination of arch arch32 and archarm, so that the same PKGBUILD can be used for x86_64 i686 or armv7h
[05:43:07] <sunshavi> that's cool
[05:44:24] <sunshavi> bill-auger: virtualbox is not there. :(
[05:44:37] <bill-auger> many will be missing from parabola though for a number of reasons
[05:45:56] <bill-auger> yes not all PKGBUILDs are there - virtualbox is not, because it requires a non-free kernel blob
[05:46:40] <bill-auger> QEMU works excellent though - no one misses virtualbox
[05:47:47] <sunshavi> second time I am being told that. I need to install freebsd on VBox. for trying a tool corchet for creating an img and booting freebsd on the orange pi
[05:47:51] <bill-auger> if you need someting in particular , youd need to get the arch PKGBUID, and the arch32 changes and merge them
[05:48:13] <sunshavi> Yes probably those are the one I need.
[05:48:25] <bill-auger> QEMU can boot anything that virtualbox can
[05:49:21] <bill-auger> that is necessarily true because they emulate the same hardware
[05:49:21] <sunshavi> probably You are right. But I would need to learn Qemu. An 10 years before Qemu was very slow
[05:49:57] <bill-auger> if you have a 64bit system with KVM, it is almost as fast as running a native system
[05:50:15] <sunshavi> i have arch32 ;)
[05:53:35] <bill-auger> i suppose that would explain the difference
[05:54:15] <bill-auger> QEMU is still slow in 32bit computers - i dont know if there is anything like KVM
[05:55:33] <bill-auger> why not just use the binary package?
[05:55:33] <sunshavi> I think this is was I am gonig to do. I am going to download the x86-64 pkgbuild for virtualbox and see where it leads me. But also i think the arch32 diffs should be published somewhere online and should be easy discoverable
[05:55:45] <bill-auger> they are
[05:56:06] <bill-auger> https://git.archlinux32.org
[05:56:07] <phrik> Title: packages - Archlinux32 package modifications (at git.archlinux32.org)
[05:56:37] <bill-auger> if something is missing, it is probably because the arch PKGBUILD soes not need patching
[05:57:33] <sunshavi> Right. It is there. then
[05:57:53] <sunshavi> abaumann was the one that tried compiling it last time on 20200522
[06:05:48] <buildmaster> i686/haskell-githash is broken (says nlopc46): https://archlinux32.org
[06:10:09] <buildmaster> pentium4/haskell-githash is broken (says rechenknecht): https://archlinux32.org
[06:14:53] <buildmaster> i686/haskell-sbv is broken (says nlopc46): https://archlinux32.org
[06:19:36] <buildmaster> pentium4/haskell-sbv is broken (says rechenknecht): https://archlinux32.org
[07:05:12] <sunshavi> bill-auger: yaourt -G did not work for virtualbox
[07:16:10] <buildmaster> i686/haskell-singletons are broken (says nlopc46): https://archlinux32.org
[07:19:51] <buildmaster> pentium4/haskell-singletons are broken (says nlopc46): https://archlinux32.org
[08:15:08] <buildmaster> pentium4/blender is broken (says nlopc46): https://archlinux32.org
[08:44:01] Alina-malina_ is now known as Alina-malina
[08:45:42] <buildmaster> i686/blender is broken (says rechenknecht): https://archlinux32.org
[08:50:41] <buildmaster> i686/linux is broken (says nlopc46): https://archlinux32.org
[08:52:00] <buildmaster> pentium4/linux is broken (says nlopc46): https://archlinux32.org
[08:52:47] <buildmaster> i686/linux-lts are broken (says rechenknecht): https://archlinux32.org
[08:53:08] <buildmaster> pentium4/linux-lts are broken (says nlopc46): https://archlinux32.org
[08:55:00] <buildmaster> pentium4/linux-zen is broken (says nlopc46): https://archlinux32.org
[08:55:28] <buildmaster> i686/linux-zen is broken (says rechenknecht): https://archlinux32.org
[09:16:30] <bill-auger> yaourt is a AUR helper isnt it?
[09:16:59] <bill-auger> ive never used it; but i beleive that it can only install PKGBUILDs from AUR
[09:17:25] <bill-auger> it surely is not going to patch a PKGBUILD which you probably need to do
[09:17:42] <bill-auger> if you have the patched PKGBUILD ; just run makepkg -sri on it
[09:21:03] <girls> sunshavi: note, that "patching a PKGBUILD" is quite trivial: just append our PKGBUILD content to the upstream PKGBUILD and you're set
[09:21:25] <girls> there's also asp on arch32 or asp32 on archlinux for x86_64, which does all the necessary stuff for you
[09:22:26] <sunshavi> girls: hi. asp
[09:22:35] <sunshavi> buildmaster: wtp asp
[09:22:35] <buildmaster> sunshavi: any/asp: i486/extra (0.5.1-1.2), pentium4/extra (0.5.1-1.2), i686/extra (0.5.1-1.2)
[09:23:37] <sunshavi> I am going to turn on the arch32 hw and see how it works
[09:23:47] <girls> in case, you want to compile on x86_64 host, you find asp in our [releng] repo
[09:24:05] <girls> https://mirror.archlinux32.org
[09:24:06] <sunshavi> just arch32
[09:24:06] <phrik> Title: Index of /x86_64/releng/ (at mirror.archlinux32.org)
[09:55:52] -!- yans has joined #archlinux32
[10:06:27] <buildmaster> i686/haskell-hslua is broken (says rechenknecht): https://archlinux32.org
[10:12:59] <buildmaster> pentium4/haskell-hslua is broken (says rechenknecht): https://archlinux32.org
[10:56:16] <buildmaster> i686/haskell-binary-parser is broken (says rechenknecht): https://archlinux32.org
[11:03:00] <buildmaster> pentium4/haskell-binary-parser is broken (says rechenknecht): https://archlinux32.org
[11:09:37] <buildmaster> i686/haskell-bytestring-strict-builder is broken (says rechenknecht): https://archlinux32.org
[11:16:20] <buildmaster> pentium4/haskell-bytestring-strict-builder is broken (says rechenknecht): https://archlinux32.org
[11:22:59] <buildmaster> i686/haskell-neat-interpolation is broken (says rechenknecht): https://archlinux32.org
[11:29:50] <buildmaster> pentium4/haskell-neat-interpolation is broken (says rechenknecht): https://archlinux32.org
[11:30:04] <sunshavi> girls: I have just installed asp
[11:33:40] <bill-auger> i have been wondering how that works
[11:35:54] <bill-auger> just a suggestion; but it be way easier on me if they were standard patches, which could be applied, and diffed
[11:36:37] <bill-auger> it is quite a brain teaser to merge those changes manually
[11:36:41] <sunshavi> let's discover how it works
[11:37:24] <bill-auger> apparently, it works by manually concatenating the two files - then it "patches" in memory
[11:37:41] <bill-auger> which means you never can actually see the PKGBUILD that build the package
[11:38:10] <bill-auger> of diff it against the upstream
[11:40:22] <sunshavi> bill-auger: inspecting the PKGBUILD would be excellent
[11:40:55] <sunshavi> girls: https://termbin.com
[12:01:32] <girls> I agree, that a true patch would be more readable, but it's also harder to maintain
[12:06:39] <bill-auger> use the git-foo luke
[12:16:08] <girls> bill-auger: yes, but even that won't work automatically
[12:16:14] <girls> the current approach does
[12:16:42] <girls> but maybe we could/should consider some git-patch system
[12:16:48] <girls> with semi-automatic rebase or something
[12:20:49] <bill-auger> but blindly running evals in memory, that cant be very reliable either
[12:21:20] <bill-auger> at least with patches, you would get some clue when something wont apply
[12:21:37] <bill-auger> those evals would just fail silently, wouldnt they?
[12:22:52] <bill-auger> just an idea - archlinuxarm publishes complete working PKGBUILDs - i pretty sure that entore distros is mainly automated
[12:23:14] <girls> if the eval fails, we get a build error
[12:23:29] <girls> if the sed in the eval doesn't match, it will be silent
[12:23:53] <girls> I believe, archlinuxarm has much greater manpower than we do
[12:24:04] <bill-auger> i dont think so
[12:24:28] <bill-auger> i think its pretty much one person and a stack of build servers
[12:24:48] <girls> ok, then it's similar to us, indeed :-D
[12:26:09] <girls> if we use the git-rebase approach, *everything* will stall if some rebase fails - currently, only single packages will be affected
[12:26:14] <bill-auger> i would think that the simplest way to do it would be to use git in the normal way - git merge upstream/master and fix the conflicts
[12:26:30] <girls> I prefer rebase over merge, but: yes
[12:26:36] <bill-auger> ok that is is true
[12:27:09] <bill-auger> doing it in git would block the entire tree from upgrading until all of the conflicts were resolved
[12:27:24] <girls> we could employ git to manage the patch/diff, but apply the patch per-package outside of git
[12:27:53] <girls> so: as long as the patch still applies, the package will not be blocked
[12:28:19] <girls> can git do rebases for some sub-directories only?
[12:28:27] <girls> if it can, that would be great ...
[12:28:57] <bill-auger> sure it could - if i understand the intention
[12:29:16] <bill-auger> you want to rebase patches, one directory at a time ?
[12:29:23] <girls> not really
[12:29:33] <girls> assume upstream has history A -> B
[12:29:40] <girls> our patch applies cleanly to A, but not to B
[12:29:50] <bill-auger> thats tivial - just keep each git commit constrained to that one dir - with ABS they normally ar
[12:29:51] <girls> because of some conflict in package1/PKGBUILD
[12:29:58] <girls> but package2/PKGBUILD applies cleanly, still
[12:30:12] <girls> ah, right, we could use multiple commits ...
[12:30:16] <girls> stupid me
[12:30:51] <girls> that sounds like a good design
[12:31:05] <girls> I'll let it ripe a little in my head, to see, how it would work
[12:31:31] <bill-auger> off-hand i think the simples appraoch wouid be a dedicated branch for each package
[12:32:49] <bill-auger> the workflow would be like: sync master wth upstream - checkout lib-foo branch - rebase lib-foo onto master
[12:33:21] <bill-auger> all changes for the lib-foo PKGBUILD would be on the lib-foo branch and nothing else
[12:33:33] <bill-auger> i think that would go smoothly
[12:33:52] <girls> it would be nice to have a master branch, which contains all patches, still
[12:34:01] <girls> but that would be a trivial merge of all patch branchse
[12:34:04] <girls> *branches
[12:34:13] <bill-auger> yes that could be automated
[12:34:51] <bill-auger> a cron script could merge everything into a uniform arch32 ABS tree for publishing
[12:35:04] <girls> the same script, that rebases all patch branches
[12:36:39] <girls> we could even combine upstream's packages and community git repository into one :-D
[12:36:51] <girls> e.g. create automatic merge commits for them
[12:36:53] <bill-auger> yep i think that could work - any branch with unresolved merge conflicts, could be ignored for that run, keeping whatever was there before - and amail could be sent telling you guy "this one need attention"
[12:37:14] <girls> yup
[12:41:00] <girls> it would still create merge conflicts on those directories
[12:41:19] <girls> we should then automatically select, whatever is in the patched version
[12:48:43] <buildmaster> i686/haskell-githash is broken (says nlopc46): https://archlinux32.org
[12:52:50] <buildmaster> pentium4/haskell-githash is broken (says nlopc46): https://archlinux32.org
[13:44:59] <buildmaster> i686/haskell-binary-parser is broken (says nlopc46): https://archlinux32.org
[13:48:49] <buildmaster> pentium4/haskell-binary-parser is broken (says nlopc46): https://archlinux32.org
[13:52:30] <buildmaster> i686/haskell-bytestring-strict-builder is broken (says nlopc46): https://archlinux32.org
[13:56:10] <buildmaster> pentium4/haskell-bytestring-strict-builder is broken (says nlopc46): https://archlinux32.org
[14:00:23] <buildmaster> pentium4/haskell-neat-interpolation is broken (says nlopc46): https://archlinux32.org
[14:00:56] <buildmaster> i686/haskell-neat-interpolation is broken (says rechenknecht): https://archlinux32.org
[14:07:43] -!- KeiraT has quit [Remote host closed the connection]
[14:08:15] <buildmaster> i686/gitea is broken (says nlopc46): https://archlinux32.org
[14:13:58] -!- KeiraT has joined #archlinux32
[14:14:54] <buildmaster> pentium4/gitea is broken (says rechenknecht): https://archlinux32.org
[14:44:47] <buildmaster> i686/mattermost is broken (says nlopc46): https://archlinux32.org
[14:57:32] <buildmaster> pentium4/mattermost is broken (says rechenknecht): https://archlinux32.org
[15:07:52] <buildmaster> i686/virtualbox-host-modules-arch is broken (says rechenknecht): https://archlinux32.org
[15:10:13] <buildmaster> pentium4/virtualbox-host-modules-arch is broken (says nlopc46): https://archlinux32.org
[17:37:59] <buildmaster> i686/python is broken (says nlopc46): https://archlinux32.org
[18:37:44] <buildmaster> pentium4/matrix-appservice-irc is broken (says nlopc46): https://archlinux32.org
[18:47:33] <buildmaster> pentium4/ell is broken (says nlopc46): https://archlinux32.org
[19:06:33] <buildmaster> i686/haskell-neat-interpolation is broken (says rechenknecht): https://archlinux32.org
[19:21:12] -!- petris has quit [Read error: Connection reset by peer]
[19:22:09] <buildmaster> i686/gnumeric is broken (says nlopc46): https://archlinux32.org
[19:22:15] -!- petris has joined #archlinux32
[19:26:50] <buildmaster> any/reflector is broken (says rechenknecht): https://archlinux32.org
[19:32:17] <buildmaster> pentium4/haskell-neat-interpolation is broken (says nlopc46): https://archlinux32.org
[20:12:17] <buildmaster> i486/python is broken (says nlopc46-i486bs1): https://archlinux32.org
[20:38:55] -!- City-busz has quit [Quit: http://quassel-irc.org - Csevegjen kényelmesen. Bárhol.]
[20:38:56] -!- elibrokeit has quit [Quit: A random quit message]
[20:39:30] -!- City-busz has joined #archlinux32
[20:39:32] -!- elibrokeit has joined #archlinux32
[21:23:24] -!- petris has quit [Remote host closed the connection]
[21:24:00] -!- petris has joined #archlinux32
[21:25:33] -!- petris has quit [Client Quit]
[21:29:22] -!- petris has joined #archlinux32
[22:05:06] <buildmaster> i486/matrix-appservice-irc is broken (says nlopc46-i486bs1): https://archlinux32.org
[22:06:10] <buildmaster> i486/matterbridge is broken (says nlopc46-i486bs0): https://archlinux32.org
[22:16:06] <buildmaster> i486/ell is broken (says nlopc46-i486bs1): https://archlinux32.org
[22:17:09] <bill-auger> has abauman mentioned anything about the libxcrypt conflict - i dont see a bug report - should i open one?
[22:18:22] <bill-auger> it is rather critical - it is required by gtk and qt via libcups; so most users are affected by it
[22:55:57] <bill-auger> not sure why, but my account on the bug tracker was deleted - i had to create it again
[23:02:28] <girls> yeah, feel free to open a bug about it
[23:10:16] <girls> I think, we need to move many packages from testing to stable, simply - but which ? :-/
[23:16:52] <bill-auger> glibc is one - i just installed blibc manually from testing, rebooted and nothing is bviously broken
[23:17:06] * girls move glibc
[23:18:52] <girls> ok, done
[23:23:48] <bill-auger> ok i found another bug
[23:23:56] <girls> hmm?
[23:24:06] <bill-auger> i noticed this on my parabola i686 VM but did not follow up on it yet
[23:24:23] <bill-auger> the latest batch of upgrades breaks LXDE
[23:25:10] <bill-auger> i just noticed the same thing on my arch32 VM - could you try it maybe to verify ?
[23:25:25] <bill-auger> lxpanel actually - the desktops starts but it almost useless without the panel
[23:25:46] <girls> sry, I don't run arch32 with graphics anywhere
[23:29:59] <girls> I can find nothing wrong with lxpanel - besides, that it won't start on my box, because it has no $DISPLAY set ;-)
[23:30:24] <bill-auger> only the GTK2 version is affected - lxpanel-gtk3 still works
[23:30:43] <girls> I just installed lxpanel and checked linking and tried to run it
[23:30:54] <girls> what package should I check instead?
[23:31:04] <bill-auger> pacman -S lxde
[23:31:38] <girls> ok, this pulls in lxpanel, which is, what I had installed ...
[23:32:32] <bill-auger> yes - that entire set will allow starting openbox/lxsession
[23:34:04] <bill-auger> i dont know if lxpanel exhibits the problem on its own - it probably does though - the LXDE parts can usually be run independently of any DE
[23:35:40] <bill-auger> it gives no errors in the console; but dmesg shows segfault in libpango-1.0.so
[23:36:23] <bill-auger> segfault is always bug so ill open a BR
[23:36:29] <girls> :-)
[23:37:16] <bill-auger> i noticed something else odd recently in the kernel logs - do you know what this means
[23:37:35] <bill-auger> BPF:hdr_len not found
[23:37:54] <girls> no idea, sry
[23:40:23] <bill-auger> yea this GTK2 situation is pretty bad
[23:41:02] <bill-auger> i started leafpad so i could copy the error message - when i tried pasting leafpad segfaulted
[23:41:22] <girls> :-/
[23:50:04] <bill-auger> ok leafpad segfault is not related to GTK - it is in libgobject
[23:59:39] <buildmaster> i686/meilisearch is broken (says nlopc46): https://archlinux32.org