#archlinux32 | Logs for 2023-07-05

[00:41:53] <ConSiGno> now lxqtarchiver doesn't work D: aaahh nooo. terminal it is...
[00:42:33] <ConSiGno> Cannot mix incompatible Qt library (5.15.10) with this library (5.15.9)
[00:59:54] <KitsuWhooa> Sounds like it needs a rebuild
[01:01:09] <ConSiGno> everything needs a rebuild :|
[01:01:15] <ConSiGno> and the pipeline is stuck
[01:31:42] <ConSiGno> I Hope this gets resolved soon...
[01:33:49] <ConSiGno> I can't play Quake 2 anymore until we unravel all this
[01:39:56] <KitsuWhooa> Hah
[01:40:04] <KitsuWhooa> You can always rebuild packages locally
[01:40:50] <ConSiGno> oh? how do I do this without doing a bunch of git stuff :V
[01:41:17] <KitsuWhooa> I do it manually so idk :p
[01:41:21] <KitsuWhooa> I've heard of asp32
[01:41:32] <ConSiGno> I will rebuild sdl2 then
[01:46:45] <bill-auger> i sugegsted how - if you stay on testing and/or staging, you may have better luck with the fancy GUI stuff
[01:47:11] <ConSiGno> which is unlike any other distro I
[01:47:11] <bill-auger> or, if you are already on testing and/or staging, try reverting to the standard releases
[01:47:14] <ConSiGno> have used lol
[01:47:15] <ConSiGno> :V
[01:47:20] <ConSiGno> oh I
[01:47:22] <bill-auger> arch32 is definitely unlike any distro
[01:47:27] <ConSiGno> i am on standard releases
[01:47:29] <bill-auger> deal with it :)
[01:47:37] <ConSiGno> i am dealing with it D:
[01:56:28] <ConSiGno> okay I'm building sdl2
[01:57:53] <ConSiGno> talk to everyone tomorrow when it finishes
[02:04:36] <ConSiGno> WAIT thisis SDL3
[02:04:41] <ConSiGno> this isnt going to work at all
[02:04:47] <ConSiGno> oh nooooo
[02:37:57] <ConSiGno> maybe it will.
[02:38:01] <ConSiGno> we shall see :3
[03:05:21] <ConSiGno> nope, still getting illegal instruction
[03:19:40] <KitsuWhooa> Oh illegal instruction means either some flags are wrong, or upstream stopped caring about 32 bit builds
[03:19:47] <KitsuWhooa> And someone has to patch it to fix to
[03:19:49] <KitsuWhooa> *it
[03:22:00] <ConSiGno> augh
[03:23:38] <ConSiGno> I miss my python and my sdl2
[04:35:23] <bill-auger> i miss my 'C' with SDL 1.2 (oh wait - i still have that)
[06:21:53] -!- buildmaster has quit [Remote host closed the connection]
[06:24:13] -!- buildmaster has joined #archlinux32
[06:29:41] -!- sunshavi has quit [Ping timeout: 260 seconds]
[14:16:38] <ConSiGno> everyone dropped SDL1.2 due to many many reasons (at least in the doom commmunity they did)
[15:21:49] <bill-auger> but it is still supported upstream via the 'compat' shim - so not everyone dropped it, and may never
[16:15:16] -!- abaumann has joined #archlinux32
[16:15:16] <buildmaster> Hi abaumann!
[16:15:16] <buildmaster> !rq abaumann
[16:15:18] <phrik> buildmaster: * abaumann killed a borg on the builmaster
[18:24:24] <abaumann> mmh. I'm trying to patch the upstream package cache for nvme-cli.
[18:24:49] <abaumann> get-package-updates is not very forgiving: exit 2 if something is wrong in a package is not good behaviour, it should just skip the package
[18:43:10] <KitsuWhooa> Ouch
[18:43:12] <KitsuWhooa> Yes
[18:48:41] <abaumann> and now get-package-updates is hanging for 30 minutes already and waiting for something..
[18:48:50] <abaumann> ..additionally I had stale lockfiles blocking things.
[18:57:23] -!- abaumann has quit [Quit: leaving]
[19:00:39] -!- abaumann has joined #archlinux32
[19:00:52] <abaumann> oupsi, killed the wrong process (irssi)
[19:00:53] <abaumann> :-)
[19:01:04] <KitsuWhooa> Nice :p
[19:01:12] <abaumann> now I kill get-package-updates twice.. deleted the lock files.. and now something runs.
[19:01:32] <abaumann> I expect the list of todos to change if get-package-updates runs through..
[19:01:49] <abaumann> that's an 'if' with capital 'I' ;-)
[19:02:31] <abaumann> I mistrust the git versions I see in get-package-updates, something is fishy also because I experienced downgrades of packages..
[19:02:53] <abaumann> ..and mod_git_revisions which are not picked up and simply stick to 000000..
[19:03:17] <abaumann> I have to do something about blacklisted packages, they still slow down things a lot..
[19:05:01] <abaumann> the new firefox is aeh... more crashing than anything else (that's on 64-bit).
[19:06:54] <abaumann> grpc ce9dbad35f81c83dd60f34eba1dcb51955b575ec 41da7ae256d15e8d2c71d5002dbc335d3d3924aa extra
[19:07:42] <abaumann> the mod_revision is the revision of the whole packages32 repo it seems not of the single commit of for instance extra/grpc/PKGBUILD
[19:07:48] <abaumann> aha
[19:11:06] <abaumann> ok, basically get-package-updates either waits on mysql or on curl, if pkginfo is cached, curl should be really fast (but it looks rather slow)
[19:12:18] <abaumann> ah, ok, it is updating pkginfos, so the curl wait time is basically makepkg --printsrcinfo running on the buildmaster
[19:12:21] <abaumann> good.
[19:16:21] <abaumann> actually, the classic approach here would be to decouple things into smaller pieces: have a git state updater which just freshes the pkginfo cache, if you do it on-demand in get-package-updates, you loose parallelism and you have to wait for caching (worst case for things you will not use again)
[19:17:16] <abaumann> mmh. actually. it tickles me to do some munin monitoring on cache metrics..
[19:17:32] <abaumann> ..doing that in the job too at the moment and it proves to be really helpful.
[19:30:46] <abaumann> nvme-cli 247bbbd1a8c09ea1bbc63ac926ceb4d70aae4b3b 0000000000000000000000000000000000000000 extra "make_source_info" did not return a "pkgbase" - eh, what?
[19:30:49] <abaumann> sigh. again.
[19:32:09] <abaumann> should have deleted old cached items :-)
[19:46:20] <KitsuWhooa> This is going well :p
[19:46:43] <abaumann> nvmi-cli 0000000000000000000000000000000000000000 41da7ae256d15e8d2c71d5002dbc335d3d3924aa extra "make_source_info" did not return a "pkgbase" - eh, what?
[19:47:02] <abaumann> not so much, now it thinks upstream disappeared and I have a local version 41da7ae256d15e8d2c71d5002dbc335d3d3924aa
[19:48:03] <abaumann> so, just patching the nvme-install file locally doesn't work. let's remove and try again.. in the end. it's faster now in retrying and failing. :-)
[19:49:26] <abaumann> people don't realize how important fast build turnaround is in software engineering..
[19:49:57] <abaumann> ..otoh it's nice to hit 'make' and the go for a coffee. ;-)
[19:50:14] <T`aZ> software guys do both :p
[19:51:01] <abaumann> lol :-)
[19:51:14] <T`aZ> definitely not gonna fix the build system for free if i cant enjoy 1h/day of idling waiting for compilation anymore
[19:54:30] <abaumann> my patched nvme-cli-247bbbd1a8c09ea1bbc63ac926ceb4d70aae4b3b.tar.gz got overwritten
[19:56:13] <abaumann> so, enough for today..
[19:56:14] -!- abaumann has quit [Quit: leaving]