#archlinux32 | Logs for 2018-10-12

Back
[00:17:58] -!- woshty has quit [Ping timeout: 245 seconds]
[00:35:42] noctambulo is now known as Chat-11004
[00:36:20] -!- bill-auger has quit [Ping timeout: 250 seconds]
[00:37:15] -!- Chat-11004 has parted #archlinux32
[00:46:29] -!- NoobAlice1 has joined #archlinux32
[00:47:16] -!- NoobAlice has quit [Ping timeout: 268 seconds]
[01:08:38] <elibrokeit> deep42thought: pacman -Sii pkgname can tell you what the syncdb "required by" linkages are
[01:09:33] <elibrokeit> You can also get a tree graph, or a dot graph if you prefer, from pactree
[01:21:13] -!- noctambulo has joined #archlinux32
[01:26:18] -!- noctambulo has quit [Quit: WeeChat 2.2]
[01:26:57] -!- noctambulo has joined #archlinux32
[03:51:21] <buildmaster> dirty! girls, my database - so dirty :-(
[03:51:21] * buildmaster goes insane.
[05:28:04] -!- bill-auger has joined #archlinux32
[05:48:20] -!- NoobAlice1 has quit [Quit: Leaving.]
[06:05:45] -!- tyzoid has quit [Ping timeout: 244 seconds]
[06:05:45] -!- girls has quit [Ping timeout: 244 seconds]
[06:08:18] -!- tyzoid has joined #archlinux32
[06:08:18] <buildmaster> Hi tyzoid!
[06:08:18] <buildmaster> !rq tyzoid
[06:08:19] <phrik> buildmaster: <tyzoid> * oblivious user is oblivious *
[06:12:16] -!- girls has joined #archlinux32
[06:12:16] <buildmaster> Hi girls!
[06:12:17] <buildmaster> !rq girls
[06:12:17] <phrik> buildmaster: <girls> I'm married, I don't cook
[06:37:26] -!- noctambulo has quit [Quit: WeeChat 2.2]
[06:40:18] * buildmaster resumes sanity.
[07:03:59] -!- noctambulo has joined #archlinux32
[07:15:08] -!- noctambulo has quit [Quit: WeeChat 2.2]
[07:54:15] -!- abaumann has joined #archlinux32
[07:54:16] <buildmaster> Hi abaumann!
[07:54:16] <buildmaster> !rq abaumann
[07:54:16] <phrik> buildmaster: <abaumann> The next step will be to distribute code to neighbour servers or use the Amazon cluster in the build script to run tests..
[08:03:18] -!- thePiGrepper has quit [Ping timeout: 252 seconds]
[08:09:35] -!- woshty has joined #archlinux32
[08:25:50] <abaumann> I'm fedup with firefox-developer-edition clogging build slaves all time..
[08:26:01] <abaumann> ..it doesn't build anyway, so I blacklisted it.
[08:57:47] -!- deep42thought has joined #archlinux32
[08:57:48] <buildmaster> Hi deep42thought!
[08:57:48] <buildmaster> !rq deep42thought
[08:57:48] <phrik> buildmaster: <deep42thought> nah, I never ran emacs - it's a nice os, but I couldn't find a suitable editor for it
[08:57:54] <deep42thought> abaumann: ok
[08:58:47] <deep42thought> elibrokeit: huh! I knew I was missing something :-)
[09:12:31] -!- thePiGrepper has joined #archlinux32
[09:17:01] <abaumann> hi deep42thought :-)
[09:18:07] <deep42thought> Hi abaumann!
[09:18:53] <abaumann> just remembered: I should make a backup of the bootstrap repo for i486 before starting to clean it.
[09:19:17] <deep42thought> :-)
[09:19:19] <deep42thought> good idea
[09:21:23] <deep42thought> thanks to lfs, I can now build vala w/o graphviz :-)
[09:22:03] <abaumann> that's cool :-)
[09:22:23] <deep42thought> hopefully, this enables us to build graphviz at some point :-D
[09:22:52] <abaumann> mmh lfs doesn't help for rust. :-(
[09:22:54] <deep42thought> graphviz -> librsvg (amongst others) -> vala -> graphviz
[09:23:14] <abaumann> rust is a piece of g*#!!*!
[09:23:20] <deep42thought> :-/
[09:23:41] <deep42thought> doesn't one usually try to avoid rust on metals?
[09:23:52] <abaumann> !grab deep42thought
[09:23:52] <phrik> abaumann: Tada!
[09:24:04] <abaumann> might be out of context though.. :-)
[09:24:13] <deep42thought> does that matter?
[09:24:39] <abaumann> no
[09:24:41] <abaumann> :-)
[09:28:30] <deep42thought> you could start the cleanup with removing the staging.db* stuff from your bootstrap-staging - this way I will not be fooled again into updating the wrong db ;-)
[09:28:59] <abaumann> oh. sorry.
[09:29:03] <abaumann> I'm cleaning up right now..
[09:29:05] <deep42thought> np
[09:29:45] <abaumann> ah. I see a vala in bootstrap/staging
[09:29:57] <deep42thought> I just pushed it there
[09:30:41] <abaumann> I think I can delete core, community in bootstrap almost completly, if not, I trigger rebuilds for the packages which are still there.
[09:30:53] <abaumann> so we should end up with a bootstrap/staging only more or less.
[09:31:08] <deep42thought> why not call it "bootstrap"?
[09:31:49] <abaumann> yeah. because bootstrap/core where the stable ones built by hand, and the staging ones where the ones now built by hand..
[09:31:59] <abaumann> ..wait, not very logical :-)
[09:31:59] <deep42thought> ah, ok
[09:32:09] <deep42thought> your decision
[09:32:20] <abaumann> no, actually, we could go with just one bootstrap, but then we have to adapt all i486 slaves.
[09:32:45] <abaumann> what about putting it to build-support for i486?
[09:33:03] <deep42thought> confuses the buildmaster :-/
[09:33:07] <abaumann> oh. ok.
[09:33:10] <abaumann> right.
[09:33:18] <deep42thought> there is still the open to-do to upload built packages into build-support
[09:33:26] <deep42thought> but otherwise this would be the perfect solution ...
[09:33:55] <abaumann> np. we can work like this for now. and bootstrapping pentium3 will be much easier, as we can use i686 for that.
[09:33:57] <deep42thought> I think adapting all i486 build slaves should be rather easy - it's only three of them, right?
[09:34:02] <abaumann> so no bootstrapping repos necessary then.
[09:34:10] <deep42thought> yeah
[09:34:14] <abaumann> 2 of yours and one of mine, yes.
[09:35:32] <abaumann> mmh. I commented here: https://github.com
[09:35:33] <phrik> Title:Pointers · Issue #1 · archlinux-riscv/packages-cross · GitHub (at github.com)
[09:35:59] <abaumann> I think, also upstream should know about this..
[09:37:39] <abaumann> ah. I remember: I wanted to have separate staging repos for core/extra and community in order to have it easier to move packages by hand, but then I realized that staging == core/extra :-)
[09:37:51] <abaumann> that's why the setup is a little bit weird for i486 bootstraps.
[09:38:15] <deep42thought> yes, having separate repos makes total sense if you want to move the packages afterwards
[09:39:23] <abaumann> Does this bootstrap setting for pacman.conf sound reasonable (currently before all other bootstrap, later replacing them all)?
[09:39:26] <abaumann> [bootstrap]
[09:39:29] <abaumann> Server = http://archlinux32.andreasbaumann.cc
[09:39:47] <deep42thought> yes
[09:39:51] <abaumann> I will move the remaining files there after cleaning up.
[09:39:54] <abaumann> like rust, vala
[09:40:10] <abaumann> xxx-dummy
[09:40:28] <abaumann> ghostcript, docbook stuff, etc.
[09:40:47] <abaumann> ok, done.
[09:40:53] <abaumann> the bootstrap repo I mean.
[09:41:06] <abaumann> my slave didn't crumble (yet) :-)
[09:42:45] <deep42thought> yep, looks good over here, too
[09:46:13] <deep42thought> damn, yet another cycle: pango -> cairo -> librsvg -> pango
[09:46:19] <abaumann> yep.
[09:46:37] <deep42thought> lfs claims to compile pango w/p cairo, but I fail to reproduce this
[09:47:59] <abaumann> mmh. also meson-ish..
[09:48:11] <deep42thought> Dependency cairo found: NO (tried pkgconfig)
[09:48:21] <deep42thought> doesn't sound like it's not needed ...
[09:48:31] <abaumann> there is no meson option to disable it.
[09:50:49] <deep42thought> --auto_features=disabled doesn't help
[09:56:17] -!- Cthulu201 has quit [Ping timeout: 252 seconds]
[10:02:00] -!- Cthulu201 has joined #archlinux32
[10:06:42] <deep42thought> abaumann: the plan is to put the stuff in ~/bootstrap/i486, right?
[10:11:17] <abaumann> yes.
[10:11:23] <abaumann> and the DB file is bootstrab.db.tar.gz
[10:11:55] <abaumann> ah. I see a pango there.
[10:12:05] <abaumann> so, compilation worked without cairo?
[10:12:24] <deep42thought> yes
[10:12:34] <deep42thought> with my wooden hammer, I get everything to compile :-D
[10:12:45] <abaumann> :-)
[10:13:10] <deep42thought> I had to remove a few "pango"s from a few "meson.build"s, though
[10:13:45] <deep42thought> damn, it looks like the xorg-server I built yesterday cannot be installed due to missing dependencies O.o
[10:14:02] <abaumann> oi. Why do I think migration from autotoools to meson ends in a disaster..
[10:14:35] <deep42thought> because you're a pessimist with no hope - or you're a realist afraid of change ... pick your evil
[10:14:36] <abaumann> xf86-input-libinput
[10:14:52] <abaumann> !grab deep42thought
[10:14:53] <phrik> abaumann: Bazinga!
[10:15:18] <abaumann> xf86-input-libinput has quite some dependencies
[10:17:31] <deep42thought> dbus libdrm libgl libsystemd libtirpc libxext libxfixes nettle xf86-input-libinput xkeyboard-config xorg-server-common xorg-setxkbmap xorg-xauth
[10:17:55] <deep42thought> these all appear as depends of a split package of xorg-server but not as a makedepends of (my) xorg-server PKGBUILD
[10:18:16] <thePiGrepper> if I can interject here, xorg-server doesnt work at all on my netbook after 1.20, and the worst part is that the transition to meson occurs just at that point in time and screws everything up, even bisecting the tree gets very difficult </rant>
[10:18:18] <deep42thought> xf86-input-libinput is the only one which is not available, though
[10:19:04] <abaumann> xf86-input-libinput only has some mild dependencies
[10:19:19] <deep42thought> I'll take a look
[10:19:27] <deep42thought> maybe you have some idea about thePiGrepper's problem?
[10:19:30] <abaumann> libevdev libwacom libgudev mtdev
[10:19:54] <abaumann> yeah, depends on the graphic card he is using.
[10:20:37] <abaumann> let me clean the Augeas stable of packages first..
[10:20:38] <thePiGrepper> that's probably the main issue, it's one of those atom netbooks. n2800 atom cpu.
[10:21:32] <thePiGrepper> so the gpu is that one which doesnt have open source 3D acceleration support.
[10:21:46] <abaumann> mmh. Isn't that one of the GPUs, where intel decided to go full non-opensource?
[10:21:53] <thePiGrepper> it works(worked) without anything else besides xorg-server
[10:22:01] <thePiGrepper> right. that one
[10:22:02] <abaumann> https://tanguy.ortolo.eu
[10:22:04] <phrik> Title:Beware of the newest Intel Atom GPU - Tanguy Ortolo (at tanguy.ortolo.eu)
[10:22:30] <abaumann> yeah. maybe Intel changes/changed its mind. In the past they where very supportive for Linux drivers..
[10:22:42] <thePiGrepper> but, it has been working for years. been using this system with arch for like 3years.
[10:23:20] <abaumann> https://wiki.archlinux.org
[10:23:21] <phrik> Title:Intel GMA 3600 - ArchWiki (at wiki.archlinux.org)
[10:23:36] <thePiGrepper> that one, right
[10:24:01] <deep42thought> gnaaaa libinput depends on gtk3 O.o
[10:24:02] <thePiGrepper> and the issue is not 3d acceleration. I _know_ that doesnt work
[10:24:29] <abaumann> deep42thought: mmh. Gnome starts to be a big cycle universe..
[10:24:41] <deep42thought> I'll try my hammer again ...
[10:24:51] <abaumann> thePiGrepper: should actually work with intel drivers.
[10:25:06] <thePiGrepper> it's that after 1.20, kernel boots and xorg goes to a black screen, without access to ttys
[10:25:15] <abaumann> check /var/log/Xorg.0.log
[10:25:58] <thePiGrepper> abaumann: I'd need to try again(it's been a while since I just freeze xorg-server)
[10:26:41] <thePiGrepper> right now I can't reboot, but I'll try that(I think I've already done that though, but at least we'll get the log)
[10:27:04] <abaumann> https://bbs.archlinux32.org
[10:27:07] <phrik> Title:Xorg stopped working after updating yesterday / Pacman / Pacman Upgrades / Arch Linux 32 Forums (at bbs.archlinux32.org)
[10:27:34] <thePiGrepper> that's my post lol
[10:27:38] <abaumann> ah. :-)
[10:27:46] <deep42thought> lol
[10:28:37] <thePiGrepper> my main problem (I believe) is that I was just bisecting it, and there're like 100 commits which keep failing during build
[10:28:57] <thePiGrepper> and it gets really hard to do because of that.
[10:29:36] <thePiGrepper> I dont know what happens with the xorg development process but, I really think that not all commits build successfully
[10:29:38] <deep42thought> yeah, that's annoying
[10:35:29] <deep42thought> !wtf gl.h
[10:35:35] <phrik> deep42thought: extra/gst-plugins-base-libs extra/hwloc extra/libepoxy extra/mesa extra/qt4 community/cgal community/cuda community/emscripten community/fltk community/glbinding community/ntk community/ogre community/spectmorph
[10:35:43] <deep42thought> whoah
[10:35:47] <deep42thought> buildmaster: wtf gl.h
[10:35:54] <buildmaster> deep42thought: [testing] qt4 (4.8.7-25.0): /usr/share/qt4/mkspecs/common/ios/GLES/gl.h
[10:35:54] <buildmaster> [extra] gst-plugins-base-libs (1.14.2-1.0): /usr/include/gstreamer-1.0/gst/gl/gl.h
[10:35:54] <buildmaster> [extra] hwloc (1.11.10-1.0): /usr/include/hwloc/gl.h
[10:35:54] <buildmaster> [extra] libepoxy (1.5.2-1.0): /usr/include/epoxy/gl.h
[10:35:54] <buildmaster> [extra] mesa (18.2.1-1.0): /usr/include/GL/gl.h /usr/include/GLES/gl.h
[10:35:54] <buildmaster> ... (17 lines total)
[10:35:55] <thePiGrepper> one interesting thing that I realized recently is that on debian, xorg-server 1.20 _works_ on my system! I dont know why, or how they do that. maybe some default config changed?
[10:36:20] <deep42thought> thePiGrepper: did you check what patches debian applies?
[10:36:42] <deep42thought> maybe they found out what the issue is and patched it (w/o telling anyone else)
[10:37:16] <abaumann> well. the patches are promintent on the web page. you just have to find the right package :-)
[10:37:22] <abaumann> *prominent
[10:37:29] <thePiGrepper> deep42thought: not yet, I noticed that last week. I want to install that .deb on arch, that does make sense even?
[10:37:47] <deep42thought> abaumann: yes, but I have the feeling, that debian does not necessarily tell upstream about what they patch
[10:37:55] <deep42thought> but I might be wrong
[10:38:10] <deep42thought> thePiGrepper: no, that doesn't make sense
[10:38:30] <abaumann> deep42thought: ah, yes, this can be.
[10:38:32] <deep42thought> at least I'm not aware of a debian package manager for arch
[10:38:41] <abaumann> especially because upstream works on new intel-64-bit stuff only..
[10:38:46] <thePiGrepper> deep42thought: hmm, yeah I thought so
[10:39:18] <thePiGrepper> the linking dependencies are different and all, however, they _might_ be the same...
[10:39:52] <thePiGrepper> probably wont work any way. is better to just understand the build config they use
[10:39:58] <deep42thought> debian probably has older versions of _some_ libs
[10:40:02] <thePiGrepper> yeah
[10:40:21] <thePiGrepper> well, one can always use sid if that were the only issue
[10:59:25] <abaumann> I'm recreating the first i486 slave after some mild bootstrapping cleanup.. let's see. :-)
[11:00:10] <deep42thought> should be no problem, I run "rm /var/cache/pacman/pkg/*; pacman -Sy $(pacman -Qq)" from time to time w/o issues
[11:03:00] <deep42thought> how do I tell ./configure to look in /usr/include/freefont2 for some header file?
[11:19:10] <abaumann> depends on the package, sometimes you have --include-dirs or --foo-include-dirs for library foo
[11:22:33] <deep42thought> there is --includedir, but it still doesn't find the header :-(
[11:22:50] <abaumann> you can set it in the CFLAGS, CXXFLAGS maybe
[11:23:10] <deep42thought> will configure pay attention to these?
[11:23:16] <abaumann> it should
[11:23:53] <deep42thought> let's see ...
[11:24:04] <deep42thought> ... \o/ it does!
[11:24:23] <abaumann> but it might overload some other flags..
[11:24:30] <abaumann> ..like the ones from makepkg.conf
[11:24:43] <deep42thought> I just added the new include dir
[11:24:46] <deep42thought> so it should be fine
[11:30:02] <abaumann> so, I also have a working build slave again on i486 :-)
[11:30:16] <deep42thought> \o/
[11:30:25] <deep42thought> was any manual intervention necessary?
[11:31:06] <deep42thought> oh, I just noticed, one of my build slaves was not running at all O.o
[11:32:26] <abaumann> I just removed /var/lib/archbuild and all caches and evrything came up again
[11:33:18] <deep42thought> abaumann: you do not need to blacklist packages for i586 (?) if they are blacklisted for i686
[11:33:37] <deep42thought> the other way round, it's needed: e.g. blacklisting for i486 does not affect i686
[11:34:15] <deep42thought> and you should remove our modifications from the repository too, otherwise the buildmaster tries to schedule a pure-archlinux32 package ;-)
[11:35:10] <abaumann> ah. ok.
[11:35:40] <abaumann> ok. I see. :-)
[11:51:30] <deep42thought> doxygen is buildable, but fails a test in check() - should we ignore it for now?
[11:51:54] <deep42thought> ah, well it's already in staging, so we do not need to enforce a build of it
[11:52:27] <abaumann> both good options :-)
[12:20:57] <deep42thought> *grml* js52 fails with "No space left on device"
[12:21:16] <abaumann> ./configure --enable-huge-special-disk
[12:21:57] <abaumann> and my cleanup script only cleaned up half of the things to clean up
[12:44:32] <abaumann> mmh. how can I force some packages to use the newest git diff revision and not a rather old one, e.g. systemd builds against 24061dcb5b89da7464b179cb6e69ecca0ae26993, which is way old..
[13:08:30] -!- Cthulu201 has quit [Ping timeout: 268 seconds]
[13:16:24] <deep42thought> abaumann: reschedule it
[13:16:47] <deep42thought> e.g. "seed-build-list -p '^systemd$'"
[13:17:09] <deep42thought> or do you mean the git revision of the upperstream source?
[13:17:18] <abaumann> no. I meant our git revision.
[13:17:21] <abaumann> ok, will try.
[13:21:03] <abaumann> bison 2fa941462f6b1ad2e5ad2550bab95138f752e274 3c468178544cb45eb84f4832c859ece1f57994f4 core
[13:21:08] <abaumann> aha. looks better :-)
[13:21:36] <deep42thought> let's hope, I really fixed the bug which was preventing what you just did
[13:22:58] <deep42thought> abaumann: how much space do you have on your i486 build slave?
[13:23:13] <abaumann> /dev/sda3 27G 11G 16G 41% /
[13:23:25] <deep42thought> can you try to build js52 at some point?
[13:23:38] <abaumann> ok. no problem
[13:24:05] <abaumann> aeh. into bootstrap?
[13:24:14] <deep42thought> should be buildable into staging
[13:24:29] <abaumann> ah. ok. so I'll reschedule it to my slave..
[13:24:41] <deep42thought> :-)
[13:24:43] <deep42thought> thx
[13:24:46] <abaumann> np
[13:24:57] <deep42thought> btw: you got quite a todo-list for your slave ;-)
[13:25:03] <abaumann> yeah. I know.
[13:25:50] <abaumann> this is all stuff from core and a little bit extra which is in bootstrap but not regularly built.
[13:26:00] <abaumann> community I didn't even bother to compare
[13:26:08] <abaumann> so bootstrap is as clean as it can be now.
[13:26:46] <deep42thought> ok :-)
[13:26:55] <abaumann> back to rust.. and back to outdoor office.. :-)
[13:27:00] <abaumann> cu later.
[13:27:04] <deep42thought> cu
[13:27:11] -!- abaumann has quit [Quit: leaving]
[13:49:28] -!- Cthulu201 has joined #archlinux32
[13:54:17] * deep42thought wonders why all those stupid makedepends are there if you can simply remove them w/o breaking the build
[14:03:17] <thePiGrepper> deep42thought: is that the case with many/most of packages? or just a couple?
[14:03:40] <deep42thought> I cannot give hard numbers
[14:03:58] <deep42thought> but I'd say 1/3 of packages have makedepends which - if absent - does not make them fail to build
[14:04:13] <deep42thought> but - assuming upstream does it right - the outcome should be a different package, though
[14:05:10] <thePiGrepper> maybe because there are some configure options absent, but the dependencies are being left in the PKGBUILD by mistake?
[14:05:18] <deep42thought> strange, js52 still fails to compile, although, I have now 21GB of free space :-/
[14:05:37] <deep42thought> thePiGrepper: I suspect auto detection of (optional) features
[14:05:56] <deep42thought> yours myight be right, too, but who knows
[14:06:07] -!- bill-auger has quit [Ping timeout: 246 seconds]
[14:06:49] -!- bill-auger has joined #archlinux32
[14:07:31] -!- abaumann has joined #archlinux32
[14:07:32] <buildmaster> Hi abaumann!
[14:07:32] <buildmaster> !rq abaumann
[14:07:33] <phrik> buildmaster: <abaumann> not to have http://archlinux32.microsoft.com suddendly in the list of mirrors. ;-)
[14:07:42] <deep42thought> LOL
[14:07:43] <thePiGrepper> deep42thought: a couple years ago I saw something like this in the vim PKGBUILD. there was a tcl dependency but it wasnt actually being built
[14:08:07] <thePiGrepper> haha, are those msgs totally random?
[14:08:20] <deep42thought> these are actual quotes from the person joining
[14:08:38] <thePiGrepper> hahah,that's even better! LOL
[14:11:19] <deep42thought> thePiGrepper: https://wiki.archlinux.org
[14:11:20] <phrik> Title:phrik - ArchWiki (at wiki.archlinux.org)
[14:13:02] <thePiGrepper> nice, thanks
[14:18:39] -!- thePiGrepper has quit [Ping timeout: 244 seconds]
[14:23:57] -!- thePiGrepper has joined #archlinux32
[14:24:54] <thePiGrepper> I've just updated some packages and I found something curious. does it make sense if a package has 'Packager : Unknown Packager'?
[14:25:12] <deep42thought> no, it does not
[14:25:15] <deep42thought> which package?
[14:25:42] <deep42thought> current packages should not have this, but we might still have some older packages with unset packager
[14:26:46] <thePiGrepper> oh, I tried to reproduce it and it has your name now. I guess that's what happens when you run a 'pacman -Qi' during the update process..
[14:26:55] <deep42thought> :-)
[14:27:05] <deep42thought> you won't get rid of me ;-)
[14:27:41] <thePiGrepper> hehe
[14:28:33] <thePiGrepper> I've noticed that the last days have had more packages moved to stable than usual.
[14:29:45] <thePiGrepper> maybe no 'more packages' in overall amount
[14:29:54] <thePiGrepper> maybe more waves of packages, I dont know
[14:30:11] <deep42thought> i486 packages overshine everything else
[14:30:16] <deep42thought> (in the statistics)
[14:30:19] <thePiGrepper> maybe just my impression though
[14:30:31] <deep42thought> we have still 10k pending packages
[14:30:41] <deep42thought> (probably/hopefully mostly i486)
[14:30:54] <thePiGrepper> that's one thing Ive been wondering about. regarding the i486 packages
[14:31:04] <thePiGrepper> these were always there? in Arch?
[14:31:09] <deep42thought> no :-D
[14:31:21] <thePiGrepper> that's what I thought
[14:31:34] <thePiGrepper> nice
[14:31:51] <deep42thought> but they're only usable in textmode, currently
[14:31:53] <thePiGrepper> is there an specific reason for it? a forum thread or something?
[14:32:05] <thePiGrepper> xorg still unusable?
[14:32:07] <deep42thought> there is, gimme a sec
[14:32:19] <deep42thought> https://bbs.archlinux32.org
[14:32:21] <phrik> Title:i486 motivation thread / Artwork, Screenshots & Setups / Arch Linux 32 Forums (at bbs.archlinux32.org)
[14:34:49] <deep42thought> the reason, that we do not have X on i486 yet is, that it's not yet fully bootstrapped
[14:35:06] <deep42thought> although, we have xorg-server since yesterday ...
[14:35:11] <deep42thought> maybe it works with X :-)
[14:38:03] <thePiGrepper> do you mean it builds and installs just fine, but it does not work? with bootstrapped you mean that it's makedepends are not i486?
[14:38:23] <deep42thought> I mean, I didn't test it
[14:38:25] <deep42thought> I just built it
[14:38:46] <deep42thought> and with "bootstrapped" I mean, that I used my wooden hammer on the PKGBUILD in order for it to compile successfully
[14:38:55] <abaumann> actually. It works in text mode on the K6 and on the Alix. :-)
[14:39:03] <thePiGrepper> one question, regarding this. is pacstrap working with i486?
[14:39:04] <deep42thought> and X?
[14:39:06] <abaumann> no X
[14:39:28] <deep42thought> you mean pacstrap from an x86_64 host?
[14:39:28] <deep42thought> no
[14:39:31] <abaumann> but, the K6 has an MGA Matrox graphics card, I doubt there is any support for this piece of art.
[14:40:11] <deep42thought> js52 builds fine outside of systemd-nspawn
[14:40:12] <thePiGrepper> no, <our> pacstrap I guess. this wasnt forked right? is just the same old one
[14:40:17] <deep42thought> ... I'll just upload this :-/
[14:40:33] <abaumann> mmh. * abaumann is pondering if this correlates to VitrualBox, firfox, thunderbird build issues..
[14:40:33] <deep42thought> thePiGrepper: pacstrap is only a shell script IIRC
[14:40:58] * deep42thought doesn't care
[14:41:05] <deep42thought> one package less to be boot strapped :-D
[14:41:09] <abaumann> doing pacman -S base with a /mnt/etc/pacman.conf and some other stuff
[14:49:59] <thePiGrepper> abaumann: right. I forgot about that. to use it with i486, it's basically just chrooting with setarch and running it.
[14:50:26] <deep42thought> setarch won't give you i486, only i686
[14:50:45] <deep42thought> try "setarch i486 uname -m"
[14:52:15] <abaumann> You can change the Architecture on your host, set it to i486, pacstrap /somewhere.
[14:52:30] <deep42thought> you still need the uname hack
[14:52:35] <abaumann> As it is a i486 which runs on i86_64 you can even chroot into it.
[14:52:49] <abaumann> I didn't try with newest pacman versions, I had the feeling some things are different
[14:53:00] <abaumann> yeah, setarch is not wrong. :-)
[14:53:31] <deep42thought> or you could hard code the architecture to i486 in pacman.conf
[14:53:57] <abaumann> yep.
[14:58:45] <thePiGrepper> when I run setarch i686(or linux32) on x64, it goes to i686 no problem, but when running setarch i486 on my i686 system, it does not change
[14:59:02] <deep42thought> yep, that's right
[14:59:04] <abaumann> I remember something was broken in setarch.
[14:59:10] <deep42thought> it's not "broken"
[14:59:15] <deep42thought> it just can't do it
[14:59:17] <abaumann> :-)
[14:59:23] <deep42thought> it merely sets the bittiness
[14:59:38] <deep42thought> but it does not hide cpu features
[15:01:56] <deep42thought> http://man7.org
[15:01:57] <phrik> Title:setarch(8) - Linux manual page (at man7.org)
[15:02:55] <abaumann> yeah. i686 stands for IA32
[15:03:13] <thePiGrepper> if that's so, how will pacstrap pull the $arch variable from the host system?
[15:03:26] <thePiGrepper> I mean, if the system is not actually i486...
[15:04:00] <deep42thought> just replace "arch=auto" by "arch=i486" in your pacman.conf
[15:04:23] <thePiGrepper> yeah, I'll do just that
[15:05:02] <abaumann> mmh. your mirror list needs adaptions. not all packages are in the standard i486 repos.
[15:05:11] <deep42thought> ah, right :-)
[15:05:16] * deep42thought forgot that
[15:05:21] <abaumann> at least bootstrap-staging and bootstrap-core
[15:06:03] <abaumann> and bootstrap-testing :-)
[15:06:16] <abaumann> this is not production-ready, yet.
[15:07:32] <abaumann> [bootstrap]
[15:07:32] <abaumann> Server = http://archlinux32.andreasbaumann.cc
[15:07:33] <abaumann> [bootstrap-staging]
[15:07:33] <abaumann> Server = http://archlinux32.andreasbaumann.cc
[15:07:33] <abaumann> [bootstrap-community-staging]
[15:07:35] <abaumann> Server = http://archlinux32.andreasbaumann.cc
[15:07:38] <abaumann> [bootstrap-testing]
[15:07:40] <abaumann> Server = http://archlinux32.andreasbaumann.cc
[15:07:43] <abaumann> [bootstrap-community-testing]
[15:07:45] <abaumann> Server = http://archlinux32.andreasbaumann.cc
[15:07:48] <abaumann> [bootstrap-core]
[15:07:51] <abaumann> Server = http://archlinux32.andreasbaumann.cc
[15:07:53] <abaumann> [bootstrap-extra]
[15:07:56] <abaumann> Server = http://archlinux32.andreasbaumann.cc
[15:07:56] <deep42thought> !give abaumann pastebin
[15:07:57] <phrik> abaumann: Pastebin.com is swamped with advertisements and random captchas. Malware found on pastebin.com has resulted in it being blocked for some users. It injects CRLF line-endings. Please, use something else. Use something sane like https://ptpb.pw https://gist.github.com https://bpaste.net https://ix.io
[15:07:58] <abaumann> [bootstrap-community]
[15:08:01] <abaumann> Server = http://archlinux32.andreasbaumann.cc
[15:08:03] <deep42thought> oops sry
[15:08:03] <abaumann> at the end of pacman.conf
[15:08:08] <deep42thought> !give abaumann ptpb
[15:08:09] <phrik> abaumann: "<command> |& curl -F c=@- https://ptpb.pw" OR "curl -F c=@path/to/a/file https://ptpb.pw"
[15:08:14] <abaumann> lol
[15:08:15] <deep42thought> !give abaumann termbin
[15:08:17] <phrik> abaumann: "<command> |& nc termbin.com 9999"
[15:08:25] <deep42thought> that's what I was looking for
[15:08:54] <abaumann> ah yes. I'm not used to cloud-based copy-paste actions. :-)
[15:09:05] <deep42thought> hrrrmm, db-update is running like forever :-/
[15:09:14] <deep42thought> ... since 1:20h now
[15:09:25] <abaumann> oeh.
[15:09:47] <abaumann> seeding a i486 rebuild list * abaumann whistles non-guiltily
[15:10:01] <deep42thought> no, this is not due to seeding
[15:10:09] <deep42thought> this is due to too many packages in staging
[15:10:38] <abaumann> yeah. I noticed that i486 has almost everything there..
[15:10:57] <deep42thought> actually, no, it does not
[15:11:16] <abaumann> ah. that was bootstrap-staging..
[15:12:04] <deep42thought> ~800 packages in testing, ~450 in staging and ~5k total
[15:12:41] <deep42thought> i686 has ~3k packages in staging, though :-/
[15:12:46] <abaumann> ui.
[15:12:47] <deep42thought> umm, 1.8k
[15:12:48] <deep42thought> sry
[15:12:58] <deep42thought> I forgot, that I also counted signatures
[15:14:10] <abaumann> I wonder whether a i686 rust would run on i486..
[15:14:19] <deep42thought> :-D
[15:14:30] <deep42thought> that's your degree of desperation?
[15:14:48] <abaumann> when people say "cross-compiling rust" they actually mean "cross-compiling some project with rust", not the compiler/cargo itself.
[15:14:53] <abaumann> yes.
[15:15:06] <deep42thought> well, isn't rust also a "project"?
[15:15:07] <deep42thought> :-/
[15:15:22] <abaumann> technically maybe..
[15:15:43] <abaumann> ..I have some other descriptions ready.. which I will not utter here..
[15:15:49] <deep42thought> LOL
[15:17:22] <abaumann> Illegal instruction (core dumped)
[15:17:26] <abaumann> I should have known..
[15:17:28] <deep42thought> you were faster
[15:17:43] <abaumann> ..in failing.. cool.
[15:25:35] <abaumann> x.py build. cool.
[15:25:42] <deep42thought> \o/
[15:26:17] <abaumann> YASMSBOP: yet another silly make system based on python
[15:26:34] <deep42thought> !grab abaumann
[15:26:35] <phrik> deep42thought: Tada!
[15:26:39] <deep42thought> is this a real thing?
[15:26:53] <abaumann> no. but maybe I should start writting one. :->
[15:26:58] <deep42thought> :-D
[15:27:39] <thePiGrepper> well, I just tried it. there are 'some' dependencies left unresolved. hehe.
[15:27:58] <abaumann> like?
[15:29:00] <thePiGrepper> i could install man-pages successfully ;-)
[15:29:10] <abaumann> oh. congrats. :-)
[15:29:26] <abaumann> aeh, you also need staging, testing, community-staging, community-.. etc.
[15:29:34] <abaumann> let me try this termbin thingy..
[15:30:24] <abaumann> *hrmpf*, no netcat yet on i486
[15:30:33] <deep42thought> but curl
[15:30:35] <deep42thought> !ptpb
[15:30:35] <phrik> "<command> |& curl -F c=@- https://ptpb.pw" OR "curl -F c=@path/to/a/file https://ptpb.pw"
[15:30:37] <abaumann> hah.
[15:31:17] <abaumann> https://ptpb.pw
[15:31:21] <abaumann> thats pacman.conf
[15:31:35] <abaumann> https://ptpb.pw
[15:31:38] <abaumann> that's the mirrorlist
[15:31:57] <abaumann> bootstrap is on http://archlinux32.andreasbaumann.cc only for i486.
[15:31:58] <phrik> Title:Index of / (at archlinux32.andreasbaumann.cc)
[15:32:56] <abaumann> mmh. who deletes those ptpb.pw thingies?
[15:33:12] <abaumann> I should not use them for password management manybe? ;-)
[15:33:14] <deep42thought> good question
[15:33:18] <deep42thought> :-D
[15:34:00] <thePiGrepper> I think they get deleted after a certain time. yeah, bad idea lol
[15:40:55] <thePiGrepper> downloading successfully
[15:42:28] <thePiGrepper> it skips a couple packages though(including pacman). is that ok?
[15:42:39] <deep42thought> yeah
[15:44:06] <deep42thought> abaumann: please do not run a series of seed-build-list commands but instead one single command with a lot of -p options
[15:44:54] <abaumann> ui. sorry. scripting typo.
[15:44:55] <deep42thought> there is some time intensive preparation and cleanup which needs to be done only once per run
[15:45:26] <abaumann> oh. that's why it takes so long.
[15:45:32] <abaumann> I should not abort it, I reckon.
[15:46:36] <deep42thought> you're rescheduling packages that are already on the build list, too?
[15:47:58] <deep42thought> e.g. make
[15:48:52] <abaumann> aeh. yeah. ls | printf|seed was not the best of my ideas today
[15:49:41] <deep42thought> indeed
[15:50:14] <abaumann> sorry :-)
[15:50:44] <abaumann> 8/30 to be rescheduled.
[15:51:03] <deep42thought> now 8/8
[15:51:30] <deep42thought> I renamed seed-build-list, so your script will abort after the current seeding
[15:51:57] <deep42thought> \o/ texlive built successfully w/o modifications :-D
[15:52:05] <deep42thought> texlive-bin, that is
[15:52:16] <abaumann> ah. that's cool.
[15:52:32] <abaumann> so we can finally build some really crutual documentation :-)
[15:52:38] <deep42thought> :-D
[15:53:02] <deep42thought> we can finally skip all that useless removing of useless documentation
[15:53:51] <deep42thought> so, seed-build-list is at its rightful place again
[16:05:19] <abaumann> yeah. much faster if executed correctly :-)
[16:05:24] <deep42thought> :-)
[16:06:15] <deep42thought> how's rust coming along?
[16:06:21] <abaumann> painfully
[16:23:51] <abaumann> I really hate that. building rust and it downloads thousands of binary blobs onto your machine..
[16:24:54] <deep42thought> isn't that what should distinguish between open source and non-open source software?
[16:26:37] <abaumann> to be fair: it has a configure script and a config.toml, almost everything should be compilable this way. They just provide a fast way for the everyday rust user, which obviously doesn't go over bootstrapping rust. :-)
[16:27:18] <deep42thought> I have to leave now, cu later!
[16:27:32] <abaumann> ok, cu.
[16:27:38] -!- deep42thought has quit [Quit: Leaving.]
[16:38:53] -!- abaumann has quit [Quit: leaving]
[16:40:30] <elibrokeit> deep42thought, abaumann: is building rust, more exciting than building gcc? Speaking of the case where you don't have an existing version.
[16:43:49] <elibrokeit> I guess the primary difference is that gcc (IIRC) can be bootstrapped by any C++ compiler, but there's only one rust compiler at all
[16:46:52] -!- noctambulo has joined #archlinux32
[16:49:38] <elibrokeit> deep42thought, abaumann: hmm, https://bootstrapping.miraheze.org
[16:49:39] <phrik> Title:Boostrapping Specific Languages - bootstrapping (at bootstrapping.miraheze.org)
[16:49:47] <elibrokeit> leads to https://github.com
[16:49:48] <phrik> Title:GitHub - thepowersgang/mrustc: Alternative rust compiler (re-implementation) (at github.com)
[17:03:10] <thePiGrepper> apparently this would work just fine to bootstrap rust and cargo.
[17:26:04] -!- noctambulo has parted #archlinux32
[17:29:49] -!- woshty has quit [Remote host closed the connection]
[17:30:36] -!- thePiGrepper has quit [Quit: leaving]
[19:24:46] -!- abaumann has joined #archlinux32
[19:24:47] <buildmaster> Hi abaumann!
[19:24:47] <buildmaster> !rq abaumann
[19:24:48] <phrik> buildmaster: <abaumann> huh. then I suddenly have to finish and maintain it. ;-)
[19:25:05] <abaumann> elibrokeit: thanks for the rust bootstrap link, will try that on the i486 slave.
[19:25:08] -!- abaumann has quit [Client Quit]
[20:48:45] -!- deep42thought has joined #archlinux32
[20:48:46] <buildmaster> Hi deep42thought!
[20:48:46] <buildmaster> !rq deep42thought
[20:48:47] <phrik> buildmaster: <deep42thought> because you're a pessimist with no hope - or you're a realist afraid of change ... pick your evil
[20:48:58] <deep42thought> great: "/build/mono/src/mono/mono/metadata/../../mono/utils/atomic.h:343: undefined reference to `__sync_val_compare_and_swap_8'"
[20:49:08] <deep42thought> why does that sound familiar?
[21:09:04] -!- noctambulo has joined #archlinux32
[21:11:12] -!- bill-auger has quit [Remote host closed the connection]
[21:15:49] -!- bill-auger has joined #archlinux32
[21:23:41] -!- thePiGrepper has joined #archlinux32
[22:11:49] -!- deep42thought has quit [Quit: Leaving.]
[23:44:53] -!- NoobAlice has joined #archlinux32