#archlinux32 | Logs for 2022-10-06

Back
[00:08:09] -!- T`aZ has quit [Ping timeout: 252 seconds]
[00:10:11] -!- T`aZ has joined #archlinux32
[01:02:04] -!- GNUtoo has quit [Ping timeout: 258 seconds]
[01:13:50] -!- GNUtoo has joined #archlinux32
[01:35:02] -!- jonathon has quit [Ping timeout: 268 seconds]
[01:36:24] -!- jonathon has joined #archlinux32
[01:40:21] -!- bill-auger has quit [Ping timeout: 260 seconds]
[01:49:25] -!- bill-auger has joined #archlinux32
[02:19:40] -!- finsternis has quit [Read error: Connection reset by peer]
[02:23:02] -!- finsternis has joined #archlinux32
[02:35:14] -!- GNUtoo has quit [Remote host closed the connection]
[02:40:31] -!- GNUtoo has joined #archlinux32
[03:12:24] -!- GNUtoo has quit [Ping timeout: 258 seconds]
[03:19:32] -!- GNUtoo has joined #archlinux32
[04:17:42] -!- ejjdhfjsu has joined #archlinux32
[05:01:26] -!- ejjdhfjsu has quit [Remote host closed the connection]
[05:01:58] -!- ejjdhfjsu has joined #archlinux32
[05:27:52] -!- ejjdhfjsu_ has joined #archlinux32
[05:31:01] -!- ejjdhfjsu has quit [Ping timeout: 268 seconds]
[05:38:57] -!- ejjdhfjsu_ has quit [Remote host closed the connection]
[05:39:21] -!- ejjdhfjsu_ has joined #archlinux32
[06:33:57] -!- ejjdhfjsu__ has joined #archlinux32
[06:36:28] -!- ejjdhfjsu_ has quit [Ping timeout: 265 seconds]
[06:58:56] -!- ejjdhfjsu__ has quit [Remote host closed the connection]
[07:40:28] -!- titus_livius has joined #archlinux32
[08:09:22] -!- abaumann has joined #archlinux32
[08:09:23] <buildmaster> Hi abaumann!
[08:09:23] <buildmaster> !rq abaumann
[08:09:23] <phrik> buildmaster: <abaumann> if fail -> try again (hard)
[08:10:05] <abaumann> I have trouble with the build system, I start to loose control over it, I can not do the simplest things like push packages between repos (https://bugs.archlinux32.org/index.php?do=details&task_id=295) or trigger a working rebuild of libreoffice..
[08:10:06] <phrik> Title: FS#295 : [qemu][seabios]: file conflict (at bugs.archlinux32.org)
[08:11:05] -!- abaumann has quit [Client Quit]
[08:18:26] -!- abaumann has joined #archlinux32
[08:18:26] <buildmaster> Hi abaumann!
[08:18:26] <buildmaster> !rq abaumann
[08:18:28] <phrik> buildmaster: <abaumann> I'm not aware of anything
[08:20:24] -!- abaumann has quit [Client Quit]
[08:38:24] -!- abaumann has joined #archlinux32
[08:38:24] <buildmaster> Hi abaumann!
[08:38:24] <buildmaster> !rq abaumann
[08:38:25] <phrik> buildmaster: <abaumann> wasn't I saying something lately about setarch being broken? ;-)
[08:38:27] <abaumann> ruby:
[08:38:30] <abaumann> ruby-etc: /usr/lib/ruby/gems/3.0.0/specifications/default/etc-1.3.0.gemspec exists in filesystem
[08:38:33] <abaumann> ruby-fcntl: /usr/lib/ruby/gems/3.0.0/specifications/default/fcntl-1.0.1.gemspec exists in filesystem
[08:38:36] <abaumann> ruby-forwardable: /usr/lib/ruby/gems/3.0.0/specifications/default/forwardable-1.3.2.gemspec exists in filesystem
[08:38:39] <abaumann> ruby-getoptlong: /usr/lib/ruby/gems/3.0.0/specifications/default/getoptlong-0.1.1.gemspec exists in filesystem
[08:38:42] <abaumann> ruby-io-nonblock: /usr/lib/ruby/gems/3.0.0/specifications/default/io-nonblock-0.1.0.gemspec exists in filesystem
[08:38:45] <abaumann> ruby-mutex_m: /usr/lib/ruby/gems/3.0.0/specifications/default/mutex_m-0.1.1.gemspec exists in filesystem
[08:38:48] <abaumann> ruby-tmpdir: /usr/lib/ruby/gems/3.0.0/specifications/default/tmpdir-0.1.2.gemspec exists in filesystem
[08:38:51] <abaumann> Errors occurred, no packages were upgraded.
[08:38:53] <abaumann> this is upstream!
[08:38:59] <abaumann> so my motivation sinks here..
[08:39:05] <abaumann> ..to fix anything.
[08:39:37] <abaumann> I cannot bootstrap ruby at the moment and it seems, the dependencies are not ok either..
[08:40:04] <abaumann> I personally would revert the latest changes.
[08:40:08] -!- abaumann has quit [Client Quit]
[09:02:20] -!- buildmaster has quit [Remote host closed the connection]
[09:05:08] -!- buildmaster has joined #archlinux32
[12:30:13] -!- tkna has parted #archlinux32
[13:21:45] -!- abaumann has joined #archlinux32
[13:21:46] <buildmaster> Hi abaumann!
[13:21:46] <buildmaster> !rq abaumann
[13:21:46] <phrik> buildmaster: <abaumann> does it include a dependency to a mailer, so it can send emails about temperature and usage too? ;-)
[13:23:32] <abaumann> mmh. I could built ruby-hitimes with ruby-bootstrap in /opt/ruby-bootstrap-xxx
[13:23:35] <abaumann> fine. :-)
[13:24:23] <abaumann> the only trouble now is, that ruby packages now trigger complete wipings of slaves because the build-support hook is sort of the last thing it tries..
[13:26:42] <abaumann> nope. it does something weird..
[13:27:25] <abaumann> I fear I have to patch some build slaves to explicitely use ruby-bootstrap and nothing else..
[13:30:39] <abaumann> sigh. pentium4 and i486 build, i686 not..
[13:38:19] -!- deep42thought has joined #archlinux32
[13:38:19] <buildmaster> Hi deep42thought!
[13:38:19] <buildmaster> !rq deep42thought
[13:38:20] <phrik> buildmaster: <deep42thought> *mimimi* tyzoid *mimimimi*
[13:38:25] <deep42thought> hi abaumann
[13:38:30] <deep42thought> should I force push some packages?
[13:40:45] <abaumann> hi deep42thought
[13:40:50] <abaumann> no, nothing builds.
[13:41:07] <abaumann> I cannot force builds to _not_ use any ruby-XXX packages and to _use_ ruby-boostrap only
[13:41:43] <abaumann> the only thing ruby is used for is for asciidoctor, which is a ruby-reimplementation of asciidoc (which is in Python).
[13:41:59] <abaumann> I consider dropping ruby and patch in asciidoc for manpage generation.
[13:42:29] <abaumann> This will make the one user unhappy using ruby on rails as his day job tool on 32-bit. :-)
[13:44:02] <deep42thought> ruby-bootstrap package looks all right to me
[13:44:10] -!- GNUtoo has quit [Remote host closed the connection]
[13:45:02] <abaumann> yeah, but PKGBUILS usually reference all kind of ruby modules..
[13:45:13] <abaumann> .. checkdepends ruby-rake, ruby-doc for documentaion etc.
[13:45:28] <abaumann> so all base modules from ruby-stdlib need to be patched for a first build.
[13:45:51] <abaumann> and as long broken ruby-* packages are in staging, they get picked instead..
[13:46:48] <deep42thought> are the ruby packages from testing and stable ok?
[13:47:08] <abaumann> maybe..
[13:47:23] <abaumann> /opt/ruby-bootstrap-3.0.4/bin/ruby: error while loading shared libraries: libruby.so.3.0: cannot open shared object file: No such file or directory
[13:47:26] <abaumann> oups,
[13:47:30] <abaumann> no ruby-bootstrap is not ok.
[13:47:35] <deep42thought> :D
[13:47:37] <abaumann> I remember, the --disable-rpath
[13:47:55] <abaumann> this is a bad idea if building shared libs outside the usual locations.. :-)
[13:48:03] <abaumann> "let's do it again, Sam" :-)
[13:49:22] <abaumann> I have bigger problems for instance with qemu, seabios..
[13:49:28] <abaumann> I simply cannot force them to stable.
[13:49:36] <abaumann> They exist in the pool, but nowhere else..
[13:49:59] <deep42thought> which one of them?
[13:50:14] <abaumann> almost all of them, but some i486.
[13:50:24] <abaumann> hard to say as qemu has so many subpackages now..
[13:50:30] <deep42thought> seabios-1.16.0-3.3-any.pkg.tar.zst is ok
[13:50:46] <deep42thought> seabios-docs-1.16.0-3.3-any.pkg.tar.zst also
[13:51:21] <abaumann> there is a file ~/qemu on the buildmaster for user master
[13:51:40] <abaumann> that's the complete list of files..
[13:52:09] <deep42thought> but they are on the buildmaster in repositories, too
[13:52:33] <abaumann> If I do a pacman -S qemu seabios qemu-arch-extra I get versions qemu-6.2.0-4.0 qemu-arch-extra-6.2.0-4.0 seabios-1.16.0-3.3
[13:52:40] <abaumann> but there are version 7.1.0
[13:53:57] <abaumann> I usually push all files with a script (via db-update)
[13:54:09] <abaumann> ~/pushall
[13:54:27] <deep42thought> the db on the mirror look ok
[13:54:33] <deep42thought> are you sure, your mirror is all right?
[13:54:42] <abaumann> mmh..
[13:54:59] <abaumann> leme check.
[13:55:03] <deep42thought> pacman -Ss qemu-audio-alsa
[13:55:03] <deep42thought> extra/qemu-audio-alsa 7.1.0-6.0
[13:55:03] <deep42thought> QEMU ALSA audio driver
[13:55:47] <deep42thought> qemu-arch-extra is not available in any 7.x version
[13:55:50] <abaumann> maybe the db files are out of sync..
[13:55:54] <deep42thought> it's only 6.1.0 on the master mirrror
[13:56:14] <deep42thought> same for qemu itself
[13:56:20] <abaumann> yep, the files are also on my mirror
[13:56:24] <deep42thought> there's qemu-6.1.0 and qemu-6.2.0
[13:57:32] <abaumann> so, qemu is the old name?
[13:57:44] <deep42thought> ah, might be, that this package was dropped
[13:57:44] <abaumann> and now maybe the master package name is different?
[13:58:02] <deep42thought> $ pacman -Ss '^qemu$'
[13:58:02] <deep42thought> extra/qemu-base 7.1.0-8
[13:58:02] <deep42thought> A basic QEMU setup for headless environments
[13:58:02] <deep42thought> extra/qemu-desktop 7.1.0-8
[13:58:02] <deep42thought> A QEMU setup for desktop environments
[13:58:04] <deep42thought> extra/qemu-full 7.1.0-8
[13:58:06] <deep42thought> A full QEMU setup
[13:58:10] <deep42thought> yep, there's no qemu package anymore :D
[13:58:24] <abaumann> so, aha.
[13:58:29] <abaumann> you have to install qemu-base or so..
[13:58:31] <deep42thought> so we should clean up a little, then
[13:58:45] <abaumann> the qemu 6 package is an old one which most likely has still something depending on it..
[13:59:08] <deep42thought> ... or not, and delete-packages made some mistake
[13:59:21] <deep42thought> I would simply force-remove it now, it most probably doesn't work anymore anyways
[14:00:30] <abaumann> no, it has a conflict with seabios
[14:00:41] <abaumann> so you cannot install it anymore
[14:00:42] <abaumann> https://bugs.archlinux32.org
[14:00:44] <phrik> Title: FS#295 : [qemu][seabios]: file conflict (at bugs.archlinux32.org)
[14:02:33] <abaumann> Cool.
[14:02:40] <abaumann> That was an easy "fix". Thanks :-)
[14:03:04] <abaumann> /opt/ruby-bootstrap-3.0.4/bin/ruby: error while loading shared libraries: libruby.so.3.0: cannot open shared object file: No such file or directory
[14:03:15] <abaumann> yeah, I have to exclicitely enable rpath..
[14:03:23] <abaumann> ..or make an ldconfig scripts for /opt/ruby-XXX
[14:06:30] <deep42thought> some other qemu packages are also still at a 6.x version
[14:06:35] <deep42thought> should I delete those, too?
[14:07:06] <abaumann> I would say so.
[14:10:37] <deep42thought> ok, done
[14:10:43] <abaumann> thanks :-
[14:10:44] <abaumann> )
[14:13:11] <abaumann> lemme test if I can still install qemu now..
[14:14:31] <abaumann> warning: cannot resolve "edk2-armvirt", a dependency of "qemu-system-aarch64"
[14:14:34] <abaumann> warning: cannot resolve "qemu-system-aarch64=7.1.0", a dependency of "qemu-full"
[14:14:37] <abaumann> ah, ok
[14:14:47] <deep42thought> ruby-bootstrap should probably also conflict with rubygems
[14:14:48] <abaumann> this is something completely different: a broken edk2 EFI
[14:14:54] <deep42thought> if it provides the /usr/bin/gem file
[14:14:55] <abaumann> mmh. true.
[14:15:09] <abaumann> qemu-base still installs.
[14:15:56] <abaumann> qemu-desktop is also fine.
[14:22:47] <abaumann> so, I was able to build a ruby-base64 with ruby-bootstrap
[14:24:27] <abaumann> local stdlib_gems=(
[14:24:46] <abaumann> I have first to build some of those ruby standard libraries to see what we have to patch..
[14:29:48] -!- GNUtoo has joined #archlinux32
[14:37:57] <deep42thought> so what's wrong with the packages in ~/qemu ?
[14:38:52] <abaumann> nothing anymore :-)
[14:38:59] <abaumann> they all got pushed to stable.
[14:39:16] <abaumann> I was just constantly trying to install the old 'qemu' package instead of 'qemu-base'
[14:39:36] <abaumann> I rebuilt them, so I had to push them, but that went fine.
[15:00:22] <abaumann> oh, cool, gems see /opt/ruby and install rake into /opt/
[15:00:56] <abaumann> local _gemdir="$(ruby -e'puts Gem.default_dir')"
[15:01:05] <abaumann> yeah, well, that's also wrong..
[15:01:54] <abaumann> or better put, if ruby is in /opt for bootstrap you cannot deduce the installtion path for the final gems from this temporary path..
[15:02:24] <deep42thought> why do you put the ruby bootstrap into /opt in the first place?
[15:03:00] <abaumann> because I need all stdlib modules (presumambly), and if I install them to their final places they conflict with ruby-XX modules..
[15:03:22] <abaumann> ..been there.. done that. :-)
[15:03:30] <deep42thought> so you would need one bootstrap package for each level of bootstrapping, then?
[15:03:48] <abaumann> I sincerly hope that one layer is enough :-)
[15:04:56] <abaumann> # local _gemdir="$(ruby -e'puts Gem.default_dir')" local _gemdir="/usr/lib/ruby/gems/3.0.0"
[15:05:06] <abaumann> yep, that works, so I have a ruby-rake now
[15:05:34] <abaumann> this would mean, I don't have to patch out rake below (and I can not, for instance for ruby-date it is pretty much reruired)
[15:05:39] <deep42thought> can you set the Gem.default_dir in the bootstrap package, maybe?
[15:05:53] <abaumann> Good thinking..
[15:05:58] <abaumann> ..I check..
[15:06:47] <abaumann> ah. most gems end up in the right place (the standard ones)
[15:06:55] <abaumann> they are built the same way without Gem.default_dir.
[15:07:14] <abaumann> It's just rake, which is not part of stdlib and assumes that ruby is installed fully and correctly.
[15:07:24] <abaumann> That would be onle one local patch for rake temporarily..
[15:07:27] <abaumann> *one*
[15:07:38] <deep42thought> :)
[15:07:44] <deep42thought> path the rake, then
[15:07:49] <deep42thought> *patch the rake
[15:08:00] <abaumann> yep :-)
[15:08:34] <abaumann> I'll trigger builds one by one manually, at least, till I'm out of the ruby-stdlib corrundrum.
[15:08:47] <abaumann> ..then we can most likely just rebuild ruby and all the modules again..
[15:08:58] <abaumann> ..at least, that's the plan. :-)
[15:11:56] <abaumann> 'gem env gemdir' another way to do the same thing :-)
[15:12:09] <abaumann> well, on the positive side: I learn something about ruby..
[15:12:25] <abaumann> looks really much the same as Python or Perl..
[15:38:06] <abaumann> https://bugs.archlinux32.org
[15:38:07] <phrik> Title: FS#296 : Fix terminal emulators (at bugs.archlinux32.org)
[15:38:11] <abaumann> More rust craziness.
[15:38:27] <abaumann> And build time is 10 times more than C++ which is 10 times more than C..
[15:38:33] <abaumann> ..that's progress, I suppose :-)
[15:39:28] -!- GNUtoo has quit [Remote host closed the connection]
[15:56:46] -!- epony has quit [Quit: QUIT]
[15:59:55] -!- GNUtoo has joined #archlinux32
[16:32:08] <abaumann> yeah, a little bit tedious, but I see the end of the tunnel.. :-)
[16:32:32] <abaumann> will move ruby-* modules to build-support/disabled/ruby after bootstrapping (for future bootstrappings to come).
[16:36:02] -!- abaumann has quit [Quit: leaving]
[17:37:45] -!- epony has joined #archlinux32
[18:34:50] -!- GNUtoo has quit [Remote host closed the connection]
[18:40:19] -!- GNUtoo has joined #archlinux32
[19:05:14] <deep42thought> honestly, I would just keep ruby-bootstrap and friends around and just comment out all the "provides" entries, so it's defused
[19:05:41] <deep42thought> not sure, if the buildmaster will choke on packages being in build-support/disabled/ruby-*
[19:05:48] <deep42thought> but you can also try your approach :)
[20:03:46] -!- abaumann has joined #archlinux32
[20:03:47] <buildmaster> Hi abaumann!
[20:03:47] <buildmaster> !rq abaumann
[20:03:48] <phrik> buildmaster: * abaumann believes in designed obsolesence in computer technology
[20:03:56] <abaumann> deep42thought: let's see :-)
[20:16:11] <abaumann> ruby-digest has a checkdepend ruby-digest, mmh.
[20:16:34] <abaumann> presumably it can only test the installed version, so it tests the version installed last time.. :-)
[20:44:04] -!- abaumann has quit [Quit: leaving]
[21:13:37] <deep42thought> this sounds wrong for a checkdepends
[21:13:48] <deep42thought> I can understand depends=($pkgname) for something like a compiler
[21:13:59] <deep42thought> ah, makedepends=($pkgname)
[22:49:06] -!- deep42thought has quit [Quit: Leaving.]