#archlinux32 | Logs for 2023-07-07
Back
[02:47:31] -!- ConSiGno has joined #archlinux32
[02:47:47] <ConSiGno> Hope everything's okay okay :(
[02:55:18] -!- GNUtoo has quit [Ping timeout: 240 seconds]
[02:57:47] -!- GNUtoo has joined #archlinux32
[04:08:26] <ConSiGno> Yeah I did some debugging, but I can
[04:08:34] <ConSiGno> i cannot run gdb
[04:08:49] <ConSiGno> i found that sdl2 is causing illegal instruction
[04:09:10] <ConSiGno> so i think some patching is in order :(
[04:22:50] -!- ConSiGno has quit [Remote host closed the connection]
[07:10:10] -!- abaumann has joined #archlinux32
[07:10:24] <abaumann> maintainance inbound on archlinux32.org and buildmaster
[07:10:48] <abaumann> I have to update mysql
[07:12:14] <abaumann> 10.11 to 11.0
[07:18:58] <buildmaster> Hi abaumann!
[07:18:58] <buildmaster> !rq abaumann
[07:18:59] <phrik> buildmaster: <abaumann> deep42thought: you cannot break cmake, it's already broken.. ;-)
[08:32:04] <abaumann> ui, get-package-updates calls a mysql in the shell for every package, that is not so efficient
[08:32:37] <abaumann> having a proper database bindings would make bash a proper language.. which is something nobody wants, I think
[08:35:00] <abaumann> and we have quite high iowait on the buildmasterwhile the script is running, so there might be some inefficiencies in the database
[08:36:42] <abaumann> I have to force build one python package after the other, with in-between get-package-updates, this is not doable at this speed
[08:37:03] <abaumann> I would also just blindly patch (orhave a build hook) for python and just rebuild all ofpython n times
[08:38:06] <abaumann> bootstrapping python packages involves ignoring check/checkdepends/optdepends and sphinx docu building. If we would have a doc() target and a docdepends, things would be easy to automatize
[08:38:52] <abaumann> as it is now, there are small differences how sphinx is called and how documentaton is packaged.
[08:40:47] <KitsuWhooa> At least it looks like some things are being built again
[08:41:33] <abaumann> that's true.
[08:42:01] <abaumann> but I see a lot of failures..
[08:45:21] <abaumann> python-build 7d9163ce2bbd69dfb390682c9bfe703a2c2e1687 4e5b06a20ec335846c1b9fffe4610057cbcfc496 community
[08:45:24] <abaumann> "make_source_info" did not return a "pkgbase" - eh, what?
[08:45:26] <abaumann> argh
[08:46:44] <abaumann> .community? seriously?
[08:47:09] <abaumann> now packages get assigned to no longer existing repos?
[08:51:10] <abaumann> -rw-r--r-- 1 http http 40580 Jul 7 10:46 python-build-0.10.0-4.1-any.pkg.tar.zst
[08:51:13] <abaumann> -rw-r--r-- 1 http http 566 Jul 7 10:46 python-build-0.10.0-4.1-any.pkg.tar.zst.sig
[08:51:16] <abaumann> -rw-r--r-- 1 http http 43915 Jul 7 10:47 python-installer-0.7.0-3.3-any.pkg.tar.zst
[08:51:19] <abaumann> -rw-r--r-- 1 http http 566 Jul 7 10:47 python-installer-0.7.0-3.3-any.pkg.tar.zst.sig
[08:51:27] <abaumann> they got built. Ok. most likely get-package-updates was enough to trigger a rebuild
[08:56:44] <abaumann> yeah, I'll just patch all the necessaryfiles and hope things work from there, my old process with forcing is gone, it seems. Too many things are broken.
[09:00:06] <abaumann> setuptools depends on jaraco-functool/text which dpeend on setuptools, nice
[09:22:46] <abaumann> -> Extracting python-flit-3.9.0.tar.gz with bsdtar
[09:22:55] <abaumann> ERROR Missing dependencies: flit_core<4,>=3.9.0
[09:23:06] <abaumann> python-flit-core-3.9.0-1.0-any.pkg.tar.zst
[09:23:19] <abaumann> so, what should that be?
[09:34:38] <abaumann> python-poetry: "Python dependency management and packaging made easy"..
[09:34:51] <abaumann> ..this means it's absolutely terrible to bootstrap. :-)
[09:39:34] <abaumann> python-flit: Simplified packaging of Python modules, so this translates into: we include every tool we can find and slap an easy interface on top' again. dependencies.
[09:54:19] -!- abaumann has quit [Quit: leaving]
[15:11:51] -!- JerryXiao has quit [Quit: Bye]
[15:13:43] -!- JerryXiao has joined #archlinux32
[22:25:31] -!- wjlafrance has joined #archlinux32