#archlinux32 | Logs for 2024-07-25

Back
[01:10:31] -!- drathir_tor has quit [Ping timeout: 260 seconds]
[01:12:47] -!- drathir_tor has joined #archlinux32
[03:06:40] -!- Cthuutloops has joined #archlinux32
[03:45:06] -!- Cthuutloops has quit [Remote host closed the connection]
[04:00:01] -!- Cthuutloops has joined #archlinux32
[06:03:59] -!- Cthuutloops has quit [Remote host closed the connection]
[06:55:03] -!- ssserpent has joined #archlinux32
[07:30:02] <KitsuWhooa> I don't know which package it is but one of the old kde5 ones removed my /lib again
[07:37:00] <KitsuWhooa> also note to self [ 405.934649] traps: RemmRDP:RS[762] trap invalid opcode ip:b1fa619a sp:a83ff870 error:0 in libfreerdp2.so.2.9.0[b1ecd000+e8000]
[07:45:45] -!- abaumann has joined #archlinux32
[07:45:45] <buildmaster> Hi abaumann!
[07:45:45] <buildmaster> !rq abaumann
[07:45:46] <phrik> buildmaster: <abaumann> the world wide web turned bad with the advent of Javascript and CSS
[07:45:48] <abaumann> KitsuWhooa: every first post to a new topic in https://bbs.archlinux32.org ("Announcements") will be on the web page. There is some caching and delay involved.
[07:45:49] <phrik> Title: Announcements / Arch Linux 32 Forum (at bbs.archlinux32.org)
[07:46:04] <KitsuWhooa> Ah, awesome. I guess I just need an account + access to it
[07:46:28] <abaumann> ah, yes. That sort of necessary :-)
[07:53:37] <KitsuWhooa> I am decomissioning my old server; once I finish and wipe it, I'll set up a temporary arch32 builder on it
[07:53:52] <abaumann> oh. cool. :-)
[07:54:52] <KitsuWhooa> until it expires and they take it down :p
[07:55:01] <KitsuWhooa> it's not much, it's an ancient opteron
[07:55:48] <abaumann> opteron were sort of fine CPUs
[07:56:09] <abaumann> a little bit power hungry maybe
[07:58:06] -!- ssserpent has quit [Ping timeout: 276 seconds]
[08:00:15] <KitsuWhooa> agreed
[08:20:13] -!- ssserpent has joined #archlinux32
[08:51:43] -!- abaumann has quit [Ping timeout: 264 seconds]
[11:04:17] -!- ssserpent has quit [Ping timeout: 248 seconds]
[11:21:04] -!- ssserpent has joined #archlinux32
[13:30:42] -!- abaumann has joined #archlinux32
[13:30:43] <buildmaster> Hi abaumann!
[13:30:43] <buildmaster> !rq abaumann
[13:30:44] <phrik> buildmaster: <abaumann> we break things - boldly - then we unbreak them again :-)
[13:31:12] <abaumann> Found the freenode irc-logs on https://mirror.archlinux32.org they were just hidden by a index.html on https://mirror.archlinux32.org and are no longer directly accessible. Everything fine. :-)
[13:31:13] <phrik> Title: Chat Logs (at mirror.archlinux32.org)
[13:38:11] <KillerWasp> I don't want to abandon the 32-bit system, but think that I have to learn and fix the complicated system within the kernel code to fix all its intentional errors.
[13:41:47] <KillerWasp> I have barely managed to understand with a lot of effort how the memory is divided, and I would have to know how everything is managed to correct the errors in the swap and oom killer, and then I would have to correct the other errors that cause the kernel to stop working with a strange colorful spot on the screen and it has little to do with the kernel panic output because it shows absolutely nothing.
[13:43:13] <T`aZ> without sounding pessimistic, unless you are getting paid for it, or have tons of free time, it's a job in itself
[13:43:13] <KillerWasp> It seems more productive to create your own kernel from scratch and recycle some modules that seem functional and necessary.
[13:44:50] <T`aZ> i also do think that picking the latest but still usable in 32bit kernel version without major kernel part crashing everywhere is the way to go, until x86 dies completely
[13:45:07] <T`aZ> i would hope debian is already doing this
[13:45:34] <abaumann> Sooner or later we have to face the fact and use older (LTS) kernels as 32-bit kernels. Most my machines run a LTS 5.15 kernel now. Latest things failing where SCSI host adapters not detecting their attached hard disks..
[13:45:53] <T`aZ> but managing the latest > 6.10 kernels when the kernel devs dont pay that much attention t oit, is delusional for a one man job
[13:45:54] <abaumann> Some machines even run kernel 5.10 (as with anything newer they just get stuck at boot)
[13:46:57] <T`aZ> ideally, picking a kernel with the nex mglru feature is *really* helping memory constrained machines (aka all of them in x86 ^^)
[13:47:04] <abaumann> We should also check, maybe just some default config switches changes and you have to enable ancient hardware in .config?
[13:47:08] <T`aZ> s/nex/new/
[13:47:50] <KitsuWhooa> my laptop runs 6.9 :p
[13:47:55] <KitsuWhooa> also, unrelated, but https://www.phoronix.com
[13:48:00] <T`aZ> whooa
[13:48:11] <T`aZ> yes, sad :(
[13:48:52] <abaumann> "will map fine to SSE2", nice.
[13:48:57] <T`aZ> what about having several lts kernels, from 4 something, where we do know it works, to your latest 6.9 where most machines fails with it, and let the user decide
[13:49:15] <abaumann> there are quite some in the AUR, so just pick one there.
[13:49:29] <KitsuWhooa> abaumann: have you reported the scsi issue upstream?
[13:49:29] <T`aZ> i mean supported by default without aur
[13:49:43] <KitsuWhooa> we can't even get a single kernel package up to date, lol
[13:49:50] <T`aZ> dont expect peaople with x86 do build a kernel :/
[13:50:04] <abaumann> KitsuWhooa: nope. neither the gdth or so. I know they will not adapt the drivers, as they literaly have no hardware anymore to test it :-)
[13:50:16] <KitsuWhooa> doesn't matter
[13:50:21] <KitsuWhooa> it should be known and tracked at the very least
[13:50:26] <abaumann> KitsuWhooa: you merged the configs of some linux packages, right?
[13:50:33] <abaumann> making maintainance a little bit easier
[13:50:45] <KitsuWhooa> abaumann: yes, but I never got around to finishing it
[13:50:51] <abaumann> ah.
[13:50:55] <KitsuWhooa> still though
[13:51:01] <KitsuWhooa> upstream keeps changing minor things in the pkgbuild constantly
[13:51:07] <KitsuWhooa> it's getting tiring
[13:51:43] <abaumann> yes, another trend lately is dropping a lot of packages on the sllightest compilation error. Then you have to dig them up in gitlab and move them to the AUR.
[13:51:56] <abaumann> The risk is always to loose the git history this way.
[13:52:09] <abaumann> I adapted some 20 packages lately or so..
[13:52:15] <KitsuWhooa> unrelated, but when can we drop the community repo
[13:52:24] <KitsuWhooa> I suspect almost nothing in there works anymore
[13:52:33] <abaumann> mmh. when the latest package got rebuild for extra/core
[13:52:35] <KitsuWhooa> nevermind, wine works
[13:52:42] <abaumann> and nothing depends on a package on community.
[13:52:47] <abaumann> so, probably never ;-)
[13:52:49] <KitsuWhooa> haskell is in community
[13:52:51] <KitsuWhooa> and haskell doesn't work
[13:53:07] <abaumann> haskell is broken since a long time.
[13:53:11] <KitsuWhooa> I know :p
[13:53:15] <KitsuWhooa> that's why I made the fake pandoc
[13:53:36] <abaumann> we could move binary packages from community to extra/core, without rebuilding them
[13:53:58] <abaumann> but this requires some tricks on the mirror and buildmaster code data
[13:54:09] <KitsuWhooa> I guess I need to go through them one by one and delete anything that's blacklisted and doesn't work anymore, and fix/rebuild everything else
[13:54:24] <abaumann> sigh, sounds like work
[13:54:28] <KitsuWhooa> a lot of it, yes
[13:55:25] <abaumann> 358 packages for pentium4, 363 for i686, 278 for i486
[13:56:06] <abaumann> some packages like 'bochs' got dropped to the AUR, so those could be removed easily from community
[13:56:28] <abaumann> to be frank: we can also just keep community forever, there is not big gain to get rid of id.
[13:56:31] <abaumann> *it
[13:57:53] <KitsuWhooa> it's most likely full of packages that don't work anymore
[13:58:08] <abaumann> yes, quite likely
[13:58:26] <abaumann> but I'm afraid of removing some lib* things which cannot easily be rebuilt.
[13:59:34] <abaumann> on first sight nothing too dramatic remains in community
[13:59:58] <abaumann> no new packages will be built into community (I hope)
[14:00:08] <abaumann> lrwxrwxrwx 1 http mirror 55 Jan 19 2024 winetricks-20240105-1.0-pentium4.pkg.tar.zst -> ../../pool/winetricks-20240105-1.0-pentium4.pkg.tar.zst
[14:00:20] <abaumann> that's the newest one
[14:00:32] <abaumann> everything else is 2023 and older
[14:00:44] <KitsuWhooa> there's stuff in there from 2023?!
[14:01:02] <abaumann> maybe so things were still stuck in the build queues..
[14:02:23] <abaumann> uff. building php74 on an armv6 is a little bit slow..
[14:02:33] <abaumann> ..I should really upgrade some raspis
[15:40:13] -!- abaumann has quit [Ping timeout: 252 seconds]
[17:40:48] -!- ssserpent has quit [Quit: WeeChat 4.3.5]
[17:57:56] -!- drathir_tor has quit [Ping timeout: 260 seconds]
[18:08:19] -!- drathir_tor has joined #archlinux32
[18:30:14] -!- balrog has quit [Quit: Bye]
[19:04:36] -!- drathir_tor has quit [Ping timeout: 260 seconds]
[19:13:53] -!- drathir_tor has joined #archlinux32
[19:40:26] -!- mvchtz has quit [Quit: WeeChat 3.5]
[19:43:21] -!- mvchtz has joined #archlinux32
[20:15:13] -!- drathir_tor has quit [Remote host closed the connection]
[20:26:43] -!- drathir_tor has joined #archlinux32
[20:35:01] -!- drathir_tor has quit [Remote host closed the connection]
[21:06:42] -!- drathir_tor has joined #archlinux32
[22:51:43] -!- T`aZ has quit [Ping timeout: 264 seconds]
[22:53:03] -!- T`aZ has joined #archlinux32