#archlinux32 | Logs for 2024-01-28

Back
[00:18:04] -!- johancb has joined #archlinux32
[01:51:11] -!- drathir_tor has quit [Remote host closed the connection]
[01:56:58] -!- drathir_tor has joined #archlinux32
[05:20:17] -!- epony has quit [Remote host closed the connection]
[05:28:13] -!- epony has joined #archlinux32
[07:57:15] -!- abaumann has joined #archlinux32
[07:57:15] <buildmaster> Hi abaumann!
[07:57:15] <buildmaster> !rq abaumann
[07:57:16] <phrik> buildmaster: <abaumann> "PAM starts to hit the fan." :-)
[07:57:41] <abaumann> libtool is held up by html2man which is hold up by gettext, which has the following pkginfo error:
[07:57:44] <abaumann> gettext 0000000000000000000000000000000000000000 0000000000000000000000000000000000000000 e6d441fe1a94f9f992cdc4d64804d430184a65ac core
[07:57:47] <abaumann> "make_source_info" did not return a "pkgbase" - eh, what?
[07:57:50] <abaumann> >==> ERROR: pkgrel is not allowed to be empty.<
[07:57:52] <abaumann> >==> ERROR: pkgver is not allowed to be empty.<
[07:57:55] <abaumann> >==> ERROR: pkgname is not allowed to be empty.<
[07:57:57] <abaumann> ><
[07:58:00] <abaumann> ><
[08:01:38] -!- abaumann has quit [Client Quit]
[13:45:01] -!- n0tiz has quit [Quit: Bye]
[13:46:45] -!- n0tiz has joined #archlinux32
[13:55:22] -!- johancb has quit [Quit: Konversation terminated!]
[17:35:46] -!- mvchtz has quit [Quit: WeeChat 3.5]
[18:01:21] -!- epony has quit [Remote host closed the connection]
[18:34:15] -!- epony has joined #archlinux32
[19:06:09] <girls> gettext is only on the build list for i486 - I'm confused
[19:20:12] <KitsuWhooa> ohai girls
[19:20:19] <KitsuWhooa> since you're around, please fix rechenknecht pretty please
[19:20:20] <girls> Hi KitsuWhooa!
[19:20:49] <KitsuWhooa> both the pgp issues and the devtools32 being too old to have broken build-support
[19:20:51] <girls> lemme see ...
[19:21:07] <KitsuWhooa> we disabled it because it was just failing on every single package
[19:21:18] <girls> yeah, I can imagine
[19:23:45] <girls> gpg should be fixed :)
[19:23:54] <KitsuWhooa> can you also update devtools32?
[19:24:00] <KitsuWhooa> that should fix the other issue
[19:24:01] <girls> I'm on it
[19:24:04] <KitsuWhooa> awesome
[19:24:17] <KitsuWhooa> (sorry, not trying to be annoying)
[19:24:23] <girls> no problem :)
[19:24:28] <girls> you aren't
[19:24:58] <girls> the issue is, that it runs on a fork of devtools for my own repo - and I didn't come to refresh it for ... well ... quite some time
[19:26:19] <KitsuWhooa> ah...
[19:27:48] <girls> but it doesn't actually need to use this anymore, so I can just install devtools32 and remove devtools-archlinuxewe :)
[19:28:17] <KitsuWhooa> I've been meaning to ask
[19:28:34] <KitsuWhooa> is there a way to get buildmaster to build a package but for testing and not staging?
[19:28:42] <girls> nope
[19:28:44] <KitsuWhooa> rip
[19:28:52] <girls> but you can force any packe to testing from staging
[19:29:00] <KitsuWhooa> it'll have dependencies on newer things though
[19:29:03] <KitsuWhooa> so it won't work
[19:29:07] <girls> ah, I see
[19:29:27] <girls> then the only way (currently) is to remove the other packages from staging
[19:29:37] <KitsuWhooa> yeah that's not happening :p
[19:29:47] <KitsuWhooa> I guess I just need to finish fixing python so that things can get going again
[19:31:28] <girls> the problem with building for testing directly, would be, that it is quite hard to determine, which dependencies are ok to take from testing and which should be taken from staging - or put differently: the package would have to be built again, once the other staging things should be moved
[19:31:45] <KitsuWhooa> yeah, I understand
[19:32:00] <KitsuWhooa> it's just that staging is mostly broken at the moment, and there are a few packages that never got rebuild that are in stable/testing and don't work
[19:32:11] <KitsuWhooa> and we can't rebuild them because they are rebuilt against staging and it won't work
[19:32:13] <girls> yeah, I understand
[19:33:31] -!- mvchtz has joined #archlinux32
[19:35:15] <girls> I'm afraid, there's not much I can do about this. The only thing, I see, would be to allow to manually "inject" packages. But that would be quite some work.
[19:37:26] <KitsuWhooa> yeah, I don't like that idea
[19:37:34] <KitsuWhooa> it should only be supported for build support
[19:38:12] <girls> if it's not too many packages, you could "overwrite" them by copying the testing packages to build-support
[19:38:26] <girls> because build-support takes precedence over stagin
[19:38:28] <girls> g
[19:39:04] <KitsuWhooa> mhm, I see
[19:39:14] <KitsuWhooa> Yeah I don't think I'll do that
[19:41:06] <girls> or you could just "hack" a build slave to use the testing-$arch-build command (but then please make sure, that a) the package is scheduled again and b) you remove the hack at some point ;-p)
[19:41:33] <girls> in other words: It is not for the buildmaster to decide, but rather for the build slave