#archlinux32 | Logs for 2021-05-30

Back
[00:00:00] <drathir_tor> Segaja: thats interesting which version reporting ?
[00:00:06] <Segaja> and the disk warning at boot is gone
[00:00:16] <Segaja> 2.04 i think it was
[00:00:27] <Segaja> and a much higher framebuffer resolution
[00:01:03] <drathir_tor> Segaja: good new grub is +_
[00:01:17] <drathir_tor> +_/++*
[00:02:01] <drathir_tor> Segaja: core/lz4 1:1.9.3-2 noticable difference pacman -Qi lz4 and downgrade to official one best bet...
[00:02:49] <Segaja> i'm not too worried about lz4
[00:02:50] <drathir_tor> Segaja: just install os-prober and grub-mkconfig -o /boot/grub/grub.cfg
[00:03:03] <Segaja> i would like to figure out how to enable os-probe in grub2
[00:03:29] <drathir_tor> Segaja: all automagically should detect... ^^
[00:03:43] <Segaja> should
[00:04:37] <drathir_tor> Segaja: it doesnt ? try pacman -S ntfs-3g
[00:05:09] <Segaja> not a valid architecture
[00:05:44] <drathir_tor> Segaja: hmmm... that interesting...
[00:05:57] <drathir_tor> Segaja: what ms version ?
[00:06:27] <Segaja> sorry, ms?
[00:06:51] <drathir_tor> or more m$ mean which windows is that ?
[00:06:58] <Segaja> ah 10
[00:07:12] <Segaja> uh no
[00:07:13] <Segaja> XP
[00:07:17] <Segaja> sorry old laptop :D
[00:07:35] <Segaja> sorry. my brain ... it is a windows 7
[00:08:15] <drathir_tor> Segaja: its still ntfs i assume... i would try mount drive and again run grub rebuild...
[00:08:33] <Segaja> hm but ntfs-3g says it doens't have a valid architecture, even so pacman -Si says i686
[00:08:44] <drathir_tor> uname -a
[00:09:05] <Segaja> i686
[00:09:13] <Segaja> but what version is ntfs-3g in your repos?
[00:10:03] <drathir_tor> Cthulu201: could You look in free time at ntfs-3g ? <3
[00:10:24] <drathir_tor> Segaja: extra/ntfs-3g 2017.3.23-5 [installed]
[00:10:41] <Segaja> i see the same
[00:10:46] <Segaja> but i can't install or reinstall it
[00:10:57] <drathir_tor> Segaja: mine from x86_64 one ofc.
[00:12:14] <Segaja> :: package ntfs-3g-2017.3.23-5.0-i686 does not have a valid architecture
[00:13:16] <Segaja> uname -m -- also lists i686
[00:13:20] <drathir_tor> Segaja: one sec i give You paste need edit with Your uuids it should temporaly boot m$
[00:13:41] <drathir_tor> Segaja: probably just package buid found ;D
[00:13:45] <Segaja> i dont' need an important fix. i can also wait
[00:13:53] <drathir_tor> buid/build*
[00:14:28] <drathir_tor> Segaja: its np just need cp paste a config...
[00:14:59] <Segaja> http://archlinux32.org that looks broken
[00:15:00] <phrik> Title: Arch Linux 32 - ntfs-3g 2017.3.23-5.0 (i686) (at archlinux32.org)
[00:15:29] <Segaja> https://git.archlinux.org line 12 looks wrong
[00:15:30] <phrik> Title: PKGBUILD « trunk - svntogit/packages.git - Git clone of the 'packages' repository (at git.archlinux.org)
[00:15:52] <Segaja> ah that link goes to archlinux.org. my bad
[00:19:54] <drathir_tor> Segaja: https://gist.github.com
[00:20:22] <Segaja> thanks
[00:21:13] <drathir_tor> Segaja: 00UUID00 replace m$ uuid from ls -hall /dev/disk/by-uuid/ and sdaX wiyh correct partition
[00:21:31] <Segaja> yeah of course :)
[00:22:15] <Segaja> again: thanks for all the help
[00:22:32] <drathir_tor> Segaja: and put in new line at /etc/grub.d/40_custom
[00:24:37] <Segaja> works thanks ;)
[00:25:02] <Segaja> interesting experience to upgradae such old system and moving to archlinux32
[00:26:20] <drathir_tor> Segaja: in case throw gpt errors You can edit like https://gist.github.com but uuid i guess have priority
[00:26:38] <drathir_tor> Segaja: yea its always fun at least for me ^^^
[00:27:00] <Segaja> to be honest i'm not sure if i will stay long on AL32
[00:27:33] <Segaja> the fact that i "found" this "spare pc" might give me the opportunity to experiment and brush up on my AL install skills
[00:28:10] <drathir_tor> Segaja: no worry at least we know its possible You can store on personal git /var/log/pacman.log ^^ it would looks funny...
[00:28:36] <drathir_tor> Segaja: [2013-10-21 04:23] [PACMAN]
[00:29:27] <drathir_tor> Segaja: still no issues ;D
[00:30:02] <Segaja> lol it starts in 2010 and 2021
[00:30:07] <Segaja> lol it starts in 2010 and ends in 2021
[00:30:37] <drathir_tor> Segaja: at least found a bug in ntfs-3g still was worth in my opinion...
[00:31:13] <Segaja> do you want to see the pacman.log? :D
[00:31:55] <drathir_tor> Segaja: and thats for ppl which try to rock You when not upgrade longer than month ^^ ;p
[00:32:27] <drathir_tor> Segaja: no its fine, You just have proof its possible ;D
[00:32:59] <Segaja> i should have had the idea earlier to use that log to figure out when the last update was
[00:33:16] <Segaja> 2016-11-24 and the next line is 2021-05-29 :D
[00:33:31] <drathir_tor> Segaja: yep thats it ^^
[00:34:27] <drathir_tor> just such small shift in upgrades... ;D
[00:35:12] <drathir_tor> Segaja: from curiosity You used prebuild pacman-static or build by Yourself ?
[00:36:07] <Segaja> prebuild
[00:36:20] <drathir_tor> Segaja: k good to know...
[00:36:33] <Segaja> downloaded from the prebuild link from the AUR package comments
[00:36:47] <Segaja> but i only used pacman-static one i moved to AL32
[00:37:59] <Segaja> and then i jumped from 2016 november to end 2018 and then to 2021
[00:39:18] <drathir_tor> Segaja: official at https://pkgbuild.com
[00:39:20] <phrik> Title: Index of /~eschwartz/repo/x86_64-extracted/ (at pkgbuild.com)
[00:39:37] <drathir_tor> i see there also https://pkgbuild.com
[00:39:38] <phrik> Title: Index of /~eschwartz/repo/i686-extracted/ (at pkgbuild.com)
[00:41:29] <Segaja> i used the i868 link from https://aur.archlinux.org
[00:41:31] <phrik> Title: AUR (en) - pacman-static (at aur.archlinux.org)
[00:42:00] <drathir_tor> Segaja: i guess it would go directly to latest as well just needed keyring fix and packages conflicts mitigate...
[00:42:55] <Segaja> yeah i guess
[00:43:17] <drathir_tor> Segaja: yep same files ;D
[00:44:00] <Segaja> if it hadn't been for the fact that the setup was i686 i would have gone without pacman-static and in small steps through the ALA
[00:45:11] <drathir_tor> Segaja: it definitely could works as well if no missing packages...
[00:45:31] <Segaja> that is something that i didn't expect to be honest
[00:46:17] <drathir_tor> Segaja: if there would be some important core ones missing in archive than try to find next available...
[00:46:21] <Segaja> IMHO if the /repos list has db files from back then, then it should have the related versions in the /packages folder
[00:46:50] <drathir_tor> Segaja: yea them should match in theory...
[00:47:52] <drathir_tor> Segaja: but You know even one corrupted at aving time and poof need search next version...
[00:48:12] <drathir_tor> aving/saving*
[00:48:28] <Segaja> yeah but the next version according to the file index on AL32A was 2 years newer :D
[00:49:08] <drathir_tor> Segaja: yea some packages have not as often updated ^^
[00:49:32] <Segaja> you mean like ntfs-3g which is from 2017 ? :D
[00:50:41] <drathir_tor> but say al isnt rock stable sounds strange when You have instalation survived multiple years...
[00:51:06] <drathir_tor> Segaja: yep ^^
[00:51:29] <Segaja> i guess it gets more unstable the older your setup gets
[00:51:35] <Segaja> and also my issues were on AL32 not AL
[00:51:43] <Segaja> AL32A not ALA
[00:51:48] <drathir_tor> Segaja: i guess even some core packages probably same old, bc there not much to change update...
[00:52:51] <drathir_tor> Segaja: yea for old versions programs slowly stops working and getting updates...
[00:53:21] <Segaja> also i'm aware of the instability that roling release distributions can have
[00:53:33] <drathir_tor> Segaja: for example i saw for rpi there is no ff build anymore...
[00:54:51] <drathir_tor> Segaja: arch have pretty solid ratio fresh vs stability...
[00:54:59] <Segaja> aye
[00:55:06] <Segaja> and i hope i can soon contribute on that ;)
[00:55:53] <drathir_tor> Segaja: im glad there is al32, bc have some old devices where was no 64bit support...
[00:56:32] <Segaja> yeha i imagine there are some devices out there ;) for me it was just lazines
[00:57:32] <drathir_tor> Segaja: at least we know for sure now and have proof its possible... ^^
[01:04:31] <Segaja> anyway thanks for the help and good night ;)
[01:45:11] <drathir_tor> Segaja: np at all glad its sucessfully updated...
[06:25:05] -!- bill-auger has joined #archlinux32
[07:03:47] -!- jonathon has quit [Remote host closed the connection]
[07:04:03] -!- jonathon has joined #archlinux32
[07:05:17] -!- eschwartz has joined #archlinux32
[08:21:46] -!- eschwartz has quit [Ping timeout: 264 seconds]
[08:23:51] -!- eschwartz has joined #archlinux32
[09:32:23] -!- kiska has joined #archlinux32
[10:26:33] -!- bill-auger has quit [Remote host closed the connection]
[11:36:15] -!- abaumann has joined #archlinux32
[11:36:15] <buildmaster> Hi abaumann!
[11:36:15] <buildmaster> !rq abaumann
[11:36:16] <phrik> buildmaster: <abaumann> doesn't ## uncomment the #? ;-)
[11:36:16] <abaumann> Segaja/drathir_tor: glad the update worked :-)
[11:42:10] <abaumann> there are holes in the archive due to some lost disks and stuff
[11:42:22] <abaumann> and the keys changed sometimes, hence the corrupt packages
[11:42:37] <abaumann> the zstd transition is also a small obstacle :-)
[11:43:15] <abaumann> deep42thought added some tipps and links on https://www.archlinux32.org which can help in such upgrades..
[11:43:17] <phrik> Title: Arch Linux 32 (at www.archlinux32.org)
[12:00:14] <Segaja> i read that page and i actually thought it was something like lost disks
[12:00:23] <Segaja> the only "issue" left is ntfs-3g
[12:17:11] <abaumann> ah, still ntfs-3g for read/write support, I suppose.
[12:17:30] <abaumann> -rw-rw-rw- 1 http http 603005 Aug 30 2020 pool/ntfs-3g-2017.3.23-5.0-pentium4.pkg.tar.zst
[12:17:41] <abaumann> urgh. yeah. that version seems a little bit old.
[12:32:03] <Segaja> and i use i686 as arch not pentium4
[12:33:14] <abaumann> pacman-conf Architecture
[12:33:26] <abaumann> uname -a is not relevant, everything is i686 for uname
[12:33:46] <abaumann> -rw-rw-rw- 1 http http 599586 Aug 30 2020 pool/ntfs-3g-2017.3.23-5.0-i686.pkg.tar.zst
[12:33:49] <abaumann> -rw-rw-rw- 1 http http 603005 Aug 30 2020 pool/ntfs-3g-2017.3.23-5.0-pentium4.pkg.tar.zst
[12:33:54] <abaumann> both architectures exists, even i486 :-)
[12:34:47] <abaumann> if you have sse2 in /proc/cpuinfo in the cpu flags then you should use pentium4, i686 has not all packages built (or working)
[12:35:01] <abaumann> even worse is i486..
[12:38:56] <Segaja> pentium4 it says then
[12:39:56] <Segaja> but the package it tries to install is ntfs-3g-2017.3.23-5.0-i686
[12:48:48] <abaumann> mmh. interesting.
[12:50:21] <Segaja> yeah but as i mentioned last night already it is more or less only a test case for me if i can do it
[12:54:45] -!- volki57 has joined #archlinux32
[13:02:12] volki57 is now known as volki57_
[13:03:11] volki57_ is now known as volki57
[13:19:15] <drathir_tor> abaumann: ;D yea a little shift in updates maded ^^
[15:03:45] -!- abaumann has quit [Quit: leaving]
[15:28:28] -!- ziria has joined #archlinux32
[16:14:04] <girls> Segaja: try `pacman -Syyu` then - the normal approach to change from i686 to pentium4 (have a look at the news, when we introduced pentium4)
[16:15:05] <girls> the dates are 2019-05-01 and 2019-05-29
[16:16:30] <girls> or were you trying to install old packages (from when we didn't have pentium4 yet)?
[16:24:18] -!- Cthulu201 has quit [Quit: WeeChat 3.1]
[16:26:54] -!- Cthulu201 has joined #archlinux32
[16:49:17] <Segaja> no i'm up to date now so i should install recent packages. i will have a look at the news
[16:54:47] -!- abaumann has joined #archlinux32
[16:54:47] <buildmaster> Hi abaumann!
[16:54:47] <buildmaster> !rq abaumann
[16:54:48] <phrik> buildmaster: <abaumann> YASMSBOP: yet another silly make system based on python
[17:04:29] <Segaja> amazing. and after the switch to pentium4 i can install ntfs-3g :D
[17:08:56] <Segaja> hm grub-mkconfig still doens't see the windows
[17:19:51] <abaumann> Some NTFS/Hibernate whatever issues?
[17:21:45] <abaumann> usually a snippet in /etc/grub.d with the correct uuid should work
[17:21:48] <abaumann> https://unix.stackexchange.com
[17:21:49] <phrik> Title: linux - grub-mkconfig can't find windows 10 - Unix & Linux Stack Exchange (at unix.stackexchange.com)
[17:36:45] -!- abaumann has quit [Quit: leaving]
[17:43:22] <Segaja> the solution is much simpler. i was missing os-prober :D
[17:43:33] <Segaja> now that it is installed it finds the windows7 when running grub-mkconfig
[17:51:13] -!- volki57 has quit [Quit: ZNC 1.8.2 - https://znc.in]
[17:53:02] -!- volki57 has joined #archlinux32
[18:22:38] -!- ytg has joined #archlinux32
[18:22:45] -!- ytg has quit [Client Quit]
[18:48:00] -!- doskoi has quit [Read error: Connection reset by peer]
[19:05:13] -!- doskoi has joined #archlinux32
[19:15:14] -!- ytg has joined #archlinux32
[19:15:23] -!- ytg has quit [Client Quit]
[19:15:29] -!- oaken-source has quit [Read error: Connection reset by peer]
[19:16:47] -!- oaken-source has joined #archlinux32
[19:37:35] -!- ziria has quit [Quit: Konversation terminated!]
[20:17:38] <drathir_tor> Segaja: cd /bin && ln -s ntfsfix fsck.ntfs
[20:18:40] <drathir_tor> Segaja: and add m$ partitions to fstab automount...
[20:19:15] <drathir_tor> Segaja: could be useful as well...
[20:28:03] <Segaja> it is already in there
[20:28:18] <Segaja> i only commented it out when ntfs-3g wasn't installed ;)
[20:55:41] <oaken-source> how many machines do you need to build archlinux32 and to keep up with upstream rebuilds?
[20:55:49] <oaken-source> or, how many do you use
[21:08:53] <girls> I use 2, sometimes 3
[21:09:29] <girls> abaumann has one which is dedicated to building only, I think, and then several, which do it "on the side" IIRC
[21:09:40] <girls> the other build slaves are all dead, now :-/
[21:11:45] <oaken-source> interesting, I would have expected there to be ... more
[21:11:54] <oaken-source> are these desktop machines or beefy server blades?
[21:12:36] <girls> one server, another beefy desktop, plus optionally another normal desktop (but all around 5-15 years old)
[21:13:07] <girls> abaumann's main build slave is rather recent (2 years old?) and specifically selected for the task, so "beefy desktop" :-D
[21:13:24] <oaken-source> okay, I can work with that. thanks :)
[21:13:25] <girls> oaken-source: the problem is really the automation and to select the proper order, etc.
[21:13:36] <girls> these machines compile 24/7
[21:13:38] <oaken-source> yeah, I've been avoiding that one :|
[21:13:48] <girls> you won't be able to do that, if you do it manually :-)
[21:13:48] <oaken-source> (this is for the riscv port)
[21:14:19] <girls> regarding riscv, it's probably best to ask the archlinuxarm guys?
[21:15:40] <oaken-source> yeah, I'm over there now
[21:17:42] <oaken-source> off the top of your head -- how long does a recent gcc build roughly take you to finish?
[21:17:55] <girls> 2h?
[21:18:02] <girls> not sure
[21:18:08] <girls> something between 1 and 4h
[21:18:16] <oaken-source> oof
[21:18:19] <oaken-source> okay, thanks
[21:18:34] <oaken-source> it's 120 hours on my beagle :D
[21:19:26] <girls> that's a conjecture?
[21:20:28] <girls> or did you measure it?
[21:20:42] <oaken-source> measured it.
[21:20:53] <oaken-source> it's probably a bit off, because the testsuite incurred some weirdness
[21:21:14] <oaken-source> the go stuff timed out, even hello world would not terminate. working on it :8
[21:21:17] <oaken-source> :)
[21:21:44] <oaken-source> but the build itself was roughly 40, and LFS estimates the test suite takes roughly twice of that on top
[21:22:32] <oaken-source> I should get better time estimates in a couple days, build is currently running
[21:24:42] <girls> how many beagle do you have?
[21:24:50] <oaken-source> one.
[21:24:56] <girls> oh
[21:25:01] <oaken-source> yeah.
[21:25:15] <girls> note, that we build on x86_64
[21:25:31] <oaken-source> even the pentium4 stuff?
[21:25:34] <girls> well, except i486, which is built in i486 vms (running on x86_64)
[21:25:41] <girls> yes, pentium4 and i686 are built on x86_64
[21:25:54] <girls> we're pondering to compile i486 there, too
[21:26:00] <oaken-source> interesting.
[21:26:02] <girls> because we often run out of memory in the vms
[21:26:20] <girls> it's a lot easier on x86 than for totally-foreign architectures
[21:26:21] <oaken-source> hm, maybe I should move to VMs too
[21:26:30] <girls> we simply do "setarch i686" and are done
[21:26:50] <girls> not sure, how arm vms compare to native arm hardware
[21:26:54] <drathir_tor> build on p4 natively for sure would closed bc of incerasing global warming significantly... ;p ^^
[21:27:11] * drathir_tor *hides*
[21:37:00] <girls> oaken-source: you could also try qemu-static-something / binfmt (sry, I forgot the real name)
[21:37:15] <girls> I think, the archlinuxarm people also compile some stuff this way
[21:39:33] <oaken-source> girls: yeah, this has lead to unexpected problems in the past
[21:39:42] <oaken-source> we've done that in parabola for our arm packages
[21:39:45] <girls> ah, right
[21:39:53] <oaken-source> but I'm hesitant to use this approach for things like glibc and gcc
[21:40:25] <oaken-source> plus, it causes a load of test suite failures that could maks real problems
[21:40:45] <oaken-source> but VMs could be an option if I get a beefier box
[21:46:26] <girls> I would definitely run some benchmarks before buying x86 hardware to build for arm
[21:51:53] <oaken-source> buying? I like your optimism :8
[21:51:55] <oaken-source> :)
[21:52:19] <oaken-source> 'finding at work' would be more accurate
[21:52:24] <girls> lol
[21:53:57] <eschwartz> I believe you're referring to qemu-user
[21:54:03] <girls> yes
[21:54:04] <girls> thanks
[21:54:27] <girls> I think, I use(d) the statically compiled one, that's why "static"
[21:55:32] <eschwartz> indeed