#archlinux32 | Logs for 2018-05-24

Back
[00:17:30] <buildmaster> python-flex is broken (says nlopc46).
[00:18:49] <buildmaster> python-flasgger is broken (says nlopc46).
[00:20:40] <buildmaster> httpbin is broken (says nlopc46).
[00:20:55] -!- musty has joined #archlinux32
[01:09:24] -!- NoobAlice has joined #archlinux32
[01:09:36] -!- musty has quit [Quit: Lost terminal]
[02:16:54] -!- davor has quit [Ping timeout: 260 seconds]
[02:21:37] -!- davor has joined #archlinux32
[03:08:48] -!- bill-auger has quit [Remote host closed the connection]
[03:12:24] -!- bill-auger has joined #archlinux32
[05:02:49] -!- belanthor has quit [Quit: Leaving]
[05:48:20] -!- NoobAlice has quit [Quit: Leaving.]
[06:16:36] -!- oaken-source has joined #archlinux32
[06:42:23] <girls> tyzoid: you mean updating the first post to reflect the changes mentioned in the last post?
[07:19:37] -!- Polichronucci_ has joined #archlinux32
[07:22:10] -!- Polichronucci has quit [Ping timeout: 260 seconds]
[07:22:10] Polichronucci_ is now known as Polichronucci
[07:58:40] -!- abaumann has joined #archlinux32
[07:58:40] <buildmaster> Hi abaumann!
[08:00:06] <abaumann> deep42thought: can you bump rust in build-support from 1.22 to 1.26? Some packages like firefox fail to build with an older rust compiler..
[08:15:15] -!- oaken-source has quit [Ping timeout: 245 seconds]
[08:22:59] <abaumann> deep42thought: mplayer is building against ffmpeg 3.4.2, but 4.0 is in testing. Why?
[08:28:23] -!- deep42thought has joined #archlinux32
[08:28:23] <buildmaster> Hi deep42thought!
[08:28:28] <deep42thought> Hi abaumann
[08:28:41] <deep42thought> we should trim down the [build-support] packages at some point, anyway :-)
[08:31:02] -!- phrik has quit [Ping timeout: 245 seconds]
[08:32:28] <abaumann> hi deep42thought.
[08:32:39] <abaumann> yes. we can also just remove rust 1.22 from build-support?
[08:32:42] -!- eschwartz has quit [Ping timeout: 245 seconds]
[08:32:42] -!- City-busz has quit [Ping timeout: 245 seconds]
[08:33:08] <deep42thought> I forgot, what it was needed for
[08:33:14] <deep42thought> I'll just remove it
[08:33:21] <deep42thought> something may fail to rebuild :-)
[08:33:24] <deep42thought> ah
[08:33:27] <deep42thought> I remember
[08:33:28] <abaumann> maybe it was used to force a certain version on an older build of firefox.
[08:33:37] <deep42thought> no, not a certain version
[08:34:04] <deep42thought> but something needed to build with rust instead of rustup (?) and thus we added it to build-support, because rustup was in testing and rust in stable
[08:34:26] <abaumann> ah.
[08:35:13] <deep42thought> regarding mplayer: is ffmpeg-4 installed or is ffmpeg-3 installen?
[08:35:17] <deep42thought> *installed
[08:35:40] <deep42thought> because: the only difference I can find is in the versions of the provided stuff
[08:35:43] <abaumann> if I have a look at archbuild, I see ffmpeg 3.4.2-2.0
[08:35:54] <abaumann> There is an ABI bump at least in libswscale.so
[08:36:07] <abaumann> So currently mplayer on testing is complaining about it.
[08:36:16] <deep42thought> so we need to rebuild it
[08:36:32] <deep42thought> that issue might be due to not-yet-having versioned dependencies in the database
[08:36:33] <abaumann> I did trigger a rebuild of mplayer, but it was using ffmpeg 3 and not 4.
[08:36:40] <deep42thought> oh
[08:37:17] <abaumann> now I triggered a rebuild of ffmpeg 4 (which was maybe not the smartest thing to do)
[08:38:46] <deep42thought> the question is, which mplayer dependency depends on a version provided by the old ffmpeg
[08:39:21] <abaumann> So mplayer needs new ffmpeg and old ffmpeg in parallel?
[08:39:40] <deep42thought> I guess, some mplayer dependency needs the old ffmpeg
[08:39:42] <deep42thought> yes
[08:39:52] <abaumann> ok. this would explain a lot.
[08:41:55] <deep42thought> abaumann: you typoed in the forum: [packages] should be [community] ;-)
[08:42:02] <abaumann> oh. :-)
[08:42:09] <deep42thought> don't worry
[08:42:18] <deep42thought> I'll rename the repository on the mirror accordingly
[08:42:26] <abaumann> urgh ;-)
[08:42:32] <deep42thought> = hot fix
[08:42:55] <abaumann> I fix it in the forum.
[08:43:13] <abaumann> might be easier. ;-)
[08:51:27] <deep42thought> how does mplayer even compile on your box?
[08:51:35] <deep42thought> it fails to apply some patch on mine O.o
[08:51:41] <abaumann> ah?
[08:51:49] <deep42thought> !ptpb
[08:51:54] <deep42thought> ah, crap
[08:53:29] <deep42thought> https://ptpb.pw
[08:53:58] <deep42thought> this might be from our modifications
[08:54:35] <abaumann> svn: E155007: None of the targets are working copies
[08:54:44] <abaumann> So it applies a patch to an empty repo?
[08:54:53] <deep42thought> oh
[08:54:57] <deep42thought> I didn't notice :-)
[08:55:00] <abaumann> :-)
[08:56:24] <deep42thought> ok
[08:56:29] <deep42thought> I found my problem
[08:56:47] <deep42thought> (I had not yet installed svn in the first run and that apparently created an empty repository)
[08:58:58] <deep42thought> if I install ffmpeg-4 manually, it gets replaced by ffmpeg-3 :-D
[09:11:58] <deep42thought> x264 depends somehow on ffmpeg-1:3.4.2
[09:12:19] <abaumann> pacman -S ffmpeg
[09:12:22] <abaumann> :: x264: installing ffmpeg (1:4.0-2.1) breaks dependency 'libavcodec.so=57-32'
[09:12:25] <abaumann> :: x264: installing ffmpeg (1:4.0-2.1) breaks dependency 'libavformat.so=57-32'
[09:12:29] <abaumann> :: x264: installing ffmpeg (1:4.0-2.1) breaks dependency 'libavutil.so=55-32'
[09:12:32] <abaumann> :: x264: installing ffmpeg (1:4.0-2.1) breaks dependency 'libswscale.so=4-32'
[09:12:38] <abaumann> that's what my build slave environment is telling me.
[09:12:53] <abaumann> so, we have to rebuild x264? and build against ffmpeg 4?
[09:13:49] <abaumann> libx264, I meant.
[09:14:03] <deep42thought> against ffmpeg 4
[09:14:06] <deep42thought> yes
[09:14:38] <abaumann> scheduling it should find ffmpeg 4 automatically, or not?
[09:14:47] <deep42thought> yes
[09:15:05] <deep42thought> except if another dependency requires ffmpeg3, too ;-)
[09:17:03] -!- oaken-source has joined #archlinux32
[09:17:15] <abaumann> no, it fetches ffmpeg 4 when building locally, so I reschedule it.
[09:17:21] <deep42thought> ok, thanks!
[09:17:53] <abaumann> np.
[09:18:10] <abaumann> I'm still trying to get my sigcontext patch into mplayer, that's what caused all the questions. :-)
[09:18:42] <deep42thought> you don't need any excuses for finding and fixing bugs ;-)
[09:20:23] <abaumann> :-)
[09:20:35] -!- phrik has joined #archlinux32
[09:20:41] <deep42thought> Hi phrik, wb!
[09:22:02] -!- eschwartz has joined #archlinux32
[09:26:02] <buildmaster> Hi phrik!
[09:34:11] <deep42thought> tyzoid: regarding wiseacre: you can also use libircclient for that
[09:38:28] -!- phrik has quit [Read error: Connection reset by peer]
[09:38:29] -!- eschwartz has quit [Quit: A random quit message]
[09:40:25] -!- phrik has joined #archlinux32
[09:40:25] <buildmaster> Hi phrik!
[09:41:18] -!- eschwartz has joined #archlinux32
[09:44:32] -!- City-busz has joined #archlinux32
[10:29:21] -!- eduardoeae has quit [Ping timeout: 248 seconds]
[11:00:15] -!- oaken-source has quit [Ping timeout: 245 seconds]
[11:02:25] -!- Polichronucci_ has joined #archlinux32
[11:04:02] -!- Polichronucci has quit [Ping timeout: 252 seconds]
[11:04:03] Polichronucci_ is now known as Polichronucci
[11:07:27] -!- Vollzornbrot has quit [Ping timeout: 240 seconds]
[11:08:02] -!- phrik has quit [Ping timeout: 265 seconds]
[11:08:56] -!- City-busz has quit [Ping timeout: 256 seconds]
[11:08:56] -!- eschwartz has quit [Ping timeout: 256 seconds]
[11:15:48] <abaumann> git.archlinux.org down?
[11:16:38] -!- eschwartz has joined #archlinux32
[11:16:49] -!- phrik has joined #archlinux32
[11:16:49] <buildmaster> Hi phrik!
[11:17:30] -!- City-busz has joined #archlinux32
[11:18:48] <deep42thought> considering how often phrik leaves and joins, I'd guess, they are working on their server(s)
[11:37:38] <abaumann> yep :-)
[12:04:14] -!- eduardoeae has joined #archlinux32
[12:17:18] <abaumann> Neither PKGBUILD nor modification of PKGBUILD found for package "libmongoc" from community (community), revisions c3d6e70fb782aecd7e88824e68cd3f0e69e60353 and 8b501654a55b02decffb511cd48ebd9f4351fb8e.
[12:17:22] <abaumann> huh?
[12:18:05] <deep42thought> umm, that looks wrong O.o
[12:22:35] <deep42thought> https://git.archlinux.org
[12:22:50] <deep42thought> upstream does not have it in git2svn, either
[12:23:03] <deep42thought> I wonder, how it ended up in our database :-/
[12:28:05] <deep42thought> I moved it to the deletion list (it has no upstream source anymore) - no Idea, why this didn't happen automatically
[12:32:48] <abaumann> ah. ok.
[12:33:15] <abaumann> is the buildmaster going for a nap? it doesn't hand out jobs anymore - just a lonely qemu is still building..
[12:42:34] <deep42thought> buildmaster: what's up?
[12:42:34] <buildmaster> up? I'm up for 1 day, 6 hours, 11 minutes, load average: 0.06, 0.19, 0.27
[12:42:39] <deep42thought> hmmm
[12:43:35] <deep42thought> no, it's rather, that Vollzornbrots buildknechte aren't working
[12:43:51] <abaumann> mmh.
[12:44:05] <deep42thought> let me start nlopc46
[12:44:11] <abaumann> aha. mine builds again after 20 minutes or so..
[12:44:48] <abaumann> maybe they got stuck while updating from the git mirrors?
[12:45:03] <deep42thought> no, I can't even log in
[12:45:12] <abaumann> ah. ok.
[12:45:21] <abaumann> so the machines dived.. :-(
[12:45:40] <abaumann> that's not even proper Germanglish.. :-)
[12:46:05] <deep42thought> they drowned
[12:46:14] <abaumann> in what? ;-)
[12:46:22] <abaumann> liquid cooling gone wrong.. ;-)
[12:46:24] <deep42thought> "TASK ERROR: volume 'Data:105/vm-105-disk-1.raw' does not exist"
[12:46:31] <abaumann> ui.
[12:46:53] <deep42thought> same for 106 and 108 :-(
[12:47:01] <deep42thought> I'll write a mail
[13:01:37] -!- Vollzornbrot has joined #archlinux32
[13:01:41] <Vollzornbrot> moin
[13:04:52] <abaumann> hi
[13:04:57] <Vollzornbrot> all okay?
[13:05:10] <Vollzornbrot> buildmaster
[13:05:15] <abaumann> yeah.
[13:05:21] <Vollzornbrot> good
[13:05:22] <abaumann> the buildmaster is fine.
[13:05:30] <abaumann> buildknechtX have trouble, it seems.
[13:05:30] <Vollzornbrot> okay und buildknecht?
[13:05:46] <abaumann> from deep43thought: ""TASK ERROR: volume 'Data:105/vm-105-disk-1.raw' does not exist"
[13:05:49] <abaumann> "
[13:05:52] <abaumann> same for 106 and 108
[13:06:00] <Vollzornbrot> i know, but i have restartet the vms
[13:06:05] <abaumann> ah. ok.
[13:06:30] -!- eduardoeae has quit [Ping timeout: 245 seconds]
[13:06:31] <abaumann> my slave eurobuild3 is building just fine.
[13:06:59] <Vollzornbrot> i have 3 buildslaves on my root xD
[13:07:28] <abaumann> yes, you build most of the packages :-)
[13:07:39] <Vollzornbrot> i know ^^
[13:07:51] <Vollzornbrot> where you come from?
[13:07:56] <abaumann> Switzerland.
[13:07:58] <abaumann> Zurich.
[13:08:04] <Vollzornbrot> Wer hats erfunden? :D
[13:08:15] <abaumann> Ricolaaaa. :-)
[13:08:18] <Vollzornbrot> :D
[13:19:12] -!- eduardoeae has joined #archlinux32
[13:23:15] <deep42thought> Vollzornbrot: what was wrong with the Buildknechte?
[13:23:58] <deep42thought> btw: thx for fixing that :-)
[13:25:01] <Vollzornbrot> np ;)
[14:47:51] <tyzoid> deep42thought: The releng thing, yes, but also just making sure things are up to date. IIRC don't we have a few builders running i686
[14:48:00] <tyzoid> it says it requires x86_64
[14:51:28] <deep42thought> ah, that
[14:51:34] <deep42thought> I never tested on i686
[14:51:42] <deep42thought> abaumann: you're building on i686, aren't you?
[14:53:06] <tyzoid> deep42thought: Also, do we have a minimum ram/disk requirement for a build slave?
[14:53:24] <deep42thought> it strongly depends on what package you build
[14:53:36] <deep42thought> firefox and thunderbird tend to need >8G hdd
[14:53:48] <deep42thought> others will build fine with a few MB
[14:53:57] <deep42thought> I'm afraid, there is no general rule
[14:54:08] <deep42thought> same will apply for ram, I guess
[15:01:36] <tyzoid> deep42thought: Do you have hsts on mirror.archlinux32.org?
[15:03:16] <deep42thought> no
[15:04:09] <tyzoid> ah
[15:04:12] <tyzoid> archlinux32.org does
[15:04:14] <tyzoid> 'strict-transport-security: max-age=15768000; includeSubDomains; preload'
[15:04:20] <deep42thought> oh
[15:04:22] <tyzoid> It's causing pool.mirror to fail https
[15:04:32] <tyzoid> (in browser, at least, should work in terminal)
[15:04:49] <deep42thought> Polichronucci: can you deactivate hsts for subdomains?
[15:05:36] <deep42thought> I think, hsts for the mirrors is useless - packages are signed anyway
[15:05:59] <deep42thought> btw: I still have no good idea how to sign the database without _decreasing_ security
[15:06:00] <tyzoid> deep42thought: more key problems
[15:06:01] <tyzoid> error: archlinux32-keyring: signature from "Erich Eckner (just to sign arch packages) <arch@eckner.net>" is marginal trust
[15:06:10] <tyzoid> this is after installing archlinux32-keyring-transition
[15:06:47] <deep42thought> btw: I updated the page in the forums
[15:09:08] <tyzoid> sweet
[15:09:13] <deep42thought> I don't get it
[15:09:21] <deep42thought> my key has 3 signatures in the transition package
[15:09:28] <deep42thought> and these three master keys are trusted
[15:09:36] <tyzoid> I'm on a new x86_64 system
[15:09:43] <tyzoid> so this is out of x86_64/releng
[15:10:10] <deep42thought> ah, is there an older version of the transition package???
[15:10:24] <tyzoid> archlinux32-keyring-transition-20170628-1-any.pkg.tar.xz is what's on that one
[15:10:32] <deep42thought> and -2 is in [core9
[15:10:34] <tyzoid> Didn't eschwartz sign a new one?
[15:10:34] <deep42thought> ]
[15:10:53] <deep42thought> /mnt/archlinux32/i686/core/archlinux32-keyring-transition-20180411-1-any.pkg.tar.xz
[15:10:56] <deep42thought> I meant
[15:11:34] <tyzoid> yeah, works with the new transition keyring
[15:11:40] <tyzoid> we just need to update that on releng
[15:11:44] <deep42thought> yes
[15:12:34] <deep42thought> can you try again?
[15:13:34] <tyzoid> my mirror doesn't have it yet, but I did a pacman -U from i686/core/archlinux32-keyring-transition, and that worked fine
[15:13:42] <deep42thought> ok
[15:13:48] <deep42thought> I just copied it over
[15:13:58] <deep42thought> hopefully I didn't forget anything :-D
[15:14:04] <tyzoid> Ok, you did the repo-add?
[15:14:10] <tyzoid> otherwise the sig won't update
[15:14:16] <deep42thought> I did
[15:14:31] <tyzoid> sounds good. I'll double check once it replicates
[15:14:34] <deep42thought> and I removed the old package
[15:14:36] <deep42thought> ok
[15:14:37] <tyzoid> did you bump the lastupdate?
[15:14:42] <deep42thought> no :-/
[15:15:06] <deep42thought> now, I did
[15:15:14] <tyzoid> Ok, I should have it any minute then
[15:17:13] <buildmaster> hedgewars is broken (says buildknecht2).
[15:21:39] <eschwartz> Whew!
[15:21:47] <deep42thought> hmm?
[15:22:35] <tyzoid> eschwartz: You can imagine this problem went unnoticed for quite some time, since we don't usually use our x86_64 repo on archlinux32 often
[15:22:40] <eschwartz> deep42thought: what exactly *did* you do, then, when I gave you the transition package?
[15:23:02] <deep42thought> eschwartz: I put it into i686/core
[15:23:20] <deep42thought> so anyone transitioning from archlinux/i686 to archlinux32 could do so as advertised
[15:24:25] <tyzoid> eschwartz: the issue was that we use the transition package for our own x86_64 systems running the build tools, since they need our signed packages to install. Makes it easier than manually adding keys
[15:24:33] <abaumann> deep42thought: I'm building on 64-bit.
[15:24:35] <tyzoid> but we didn't update the transition package in our x86_64 repo
[15:24:51] <deep42thought> abaumann: ok, thx
[15:25:01] <tyzoid> I could swear we had someone building on i686
[15:25:20] <abaumann> I'm just building the i486-stuff on i686.
[15:25:37] <abaumann> my build machine is - o wonders - a full 64-bit machine :-)
[15:25:46] <eschwartz> Oh
[15:25:51] <abaumann> :-)
[15:25:55] <tyzoid> abaumann: You referring to the vm I've got on srv0?
[15:26:01] <tyzoid> that's running i486, iirc
[15:26:09] <abaumann> yes, that's one.
[15:26:20] <abaumann> And here I also have a server doing i486 stuff.
[15:26:35] <tyzoid> No, I mean it's on an 80486 qemu machine
[15:26:47] <tyzoid> not i686
[15:26:49] <abaumann> yes, that one is.
[15:27:04] * tyzoid continues with buildslave setup
[15:33:56] <deep42thought> the hedgewars build fails due to some cmake stuff which works fine on x86_64
[15:34:02] <deep42thought> I hope, we didn't break cmake :-/
[15:37:43] <tyzoid> deep42thought: https://bbs.archlinux32.org
[15:37:45] <phrik> Title:How to set up a build slave / Building / Arch Linux 32 Forums (at bbs.archlinux32.org)
[15:37:49] <tyzoid> Writing up better steps for this
[15:38:44] <deep42thought> git should be a dependency of devtools32
[15:39:30] <tyzoid> Thanks - I saw I had it already, but wanted to make sure, since I wasn't sure if it was in my base image
[15:39:42] <tyzoid> but yea, it's in the deps
[15:40:01] <abaumann> deep42thought: you cannot break cmake, it's already broken.. ;-)
[15:40:08] <deep42thought> !grab abaumann
[15:40:08] <phrik> deep42thought: Tada!
[15:40:11] <tyzoid> lol
[15:40:25] <abaumann> ah.. what a reliev to have phrik back. :-)
[15:40:35] <deep42thought> yeah
[15:40:43] <deep42thought> !q abaumann
[15:40:44] <phrik> deep42thought: <abaumann> deep42thought: you cannot break cmake, it's already broken.. ;-)
[15:40:52] <tyzoid> !offtopic
[15:40:53] <phrik> Take it to #archlinux-offtopic
[15:40:57] <tyzoid> :/
[15:41:05] <deep42thought> sry
[15:41:24] <tyzoid> No, I was just seeing if it would say archlinux32-offtopic of archlinux-offtopic
[15:41:25] <tyzoid> xD
[15:41:29] <tyzoid> you
[15:41:32] <tyzoid> you're fine
[15:41:42] <tyzoid> damn hitting enter when trying to press apostrophe.
[15:42:12] <deep42thought> there is some content which should be more wiki-like or a static page or something
[15:42:22] <deep42thought> howtos are one part :-/
[15:46:35] <tyzoid> I don't think we need to go all the way to setting up mediawiki
[15:46:41] <tyzoid> but perhaps creating a wiki project on github?
[15:47:01] <tyzoid> or making a github-pages project for documentation?
[15:47:12] <deep42thought> that
[15:54:36] <tyzoid> damn, this is one slow git clone
[15:55:01] <tyzoid> deep42thought: is build-packages cloning the entirity of community/extra/core for upstream?
[15:55:08] <deep42thought> yes
[15:55:45] <deep42thought> https://github.com git://git.archlinux.org git://git.archlinux.org
[15:55:48] <phrik> Title:GitHub - archlinux32/packages: package customizations and pure-i686 packages (at github.com)
[15:55:49] <tyzoid> yup
[15:56:09] <tyzoid> Anyway, I'm going to let this run - I've got a meeting in 5 mins that I'm heading off to
[15:56:42] <deep42thought> I'll leave any minute now, too
[15:56:49] <deep42thought> cu later
[15:59:41] -!- deep42thought has quit [Quit: Leaving.]
[16:49:27] -!- eduardoeae has quit [Quit: eduardoeae]
[16:51:25] -!- eduardoeae has joined #archlinux32
[16:56:35] -!- deep42thought has joined #archlinux32
[16:56:35] <buildmaster> Hi deep42thought!
[16:57:18] <deep42thought> tyzoid: if one wants to separate the "bin/build-packages" from the clone of the upstream packages, sourcing "conf/default.conf" is enough to do the initial clone
[16:57:49] <deep42thought> ah: conf/default.conf is _not_ meant to be copied
[16:58:00] <deep42thought> maybe the name is misleading ...
[17:05:46] <tyzoid> ah I see
[17:06:28] <deep42thought> I should clarify that at some point by moving it to lib/set-up-variables or so
[17:06:45] <deep42thought> and creating a conf/example with defaults commented
[17:06:58] <tyzoid> still cloning, btw
[17:07:19] <deep42thought> you didn't set up an infinite loop by copying conf/default.conf over to conf/local.conf?
[17:07:30] <tyzoid> I did not copy it, no
[17:07:35] <deep42thought> good :-)
[17:07:51] <deep42thought> btw: it's worth disabling the auto-cleanup on the git repositories
[17:10:58] <deep42thought> also: we could set up mirrors of the git repositories for bootstrapping
[17:11:26] <tyzoid> Cloning into bare repository '/home/builder/builder/bin/../work/repos/packages32'...
[17:11:27] <tyzoid> The authenticity of host 'github.com (192.30.253.112)' can't be established.
[17:11:27] <tyzoid> RSA key fingerprint is SHA256:nThbg6kXUpJWGl7E1IGOCspRomTxdCARLviKw6E5SY8.
[17:11:28] <tyzoid> Are you sure you want to continue connecting (yes/no)? yes
[17:11:28] <tyzoid> Warning: Permanently added 'github.com,192.30.253.112' (RSA) to the list of known hosts.
[17:11:29] <tyzoid> git@github.com: Permission denied (publickey).
[17:11:29] <tyzoid> fatal: Could not read from remote repository.
[17:11:31] <tyzoid> Please make sure you have the correct access rights
[17:11:32] <tyzoid> and the repository exists.
[17:11:33] <tyzoid> sorry for the long paste
[17:11:51] <tyzoid> but I think this should be an https clone, not an ssh clone
[17:12:14] <deep42thought> hmmm
[17:12:20] <deep42thought> I prefer ssh
[17:12:26] <deep42thought> but I see, that this is not a good default
[17:13:48] <deep42thought> try now
[17:13:55] <tyzoid> also: $ source builder/conf/default.conf
[17:13:55] <tyzoid> -bash: /home/builder/lib/common-functions: No such file or directory
[17:14:07] <tyzoid> fails if you're not in the root of the git repo
[17:14:11] <deep42thought> oh
[17:14:15] <tyzoid> not sure if that's something we want to correct for
[17:14:24] <tyzoid> or just say "yeah, run in git repo"
[17:14:47] <deep42thought> this is only true for the source itself
[17:14:56] <deep42thought> if you call the scripts, they should work from everywhere
[17:15:03] <tyzoid> yup, works without the source by calling the script
[17:16:01] <tyzoid> problem is that build-packages will get a failure on connect because you don't have the pubkey
[17:16:12] <tyzoid> but yeah, should be fine
[17:18:06] <deep42thought> "# dirty hack to get this stuff debugable from a bash"
[17:18:28] <deep42thought> I think, we should not fix this and just say, that sourcing the scripts directly is reserved for me :-)
[17:18:53] <tyzoid> "After configuration, source 'conf/default.conf' to generate a valid working tree, ssh-key and alike."
[17:19:03] <deep42thought> yeah
[17:20:04] <tyzoid> and I would recommend having the build-packages process exit upon ssh failure to buildmaster
[17:20:13] <tyzoid> that way, you don't have to Ctrl+C on initial setup
[17:20:28] <tyzoid> and it'll allow a systemd unit to automatically restart it if there's a temporary problem
[17:20:33] <deep42thought> it exits if the exit code is >1
[17:20:46] <tyzoid> ERROR: Unknown exit code 255 from 'get-assignment'.
[17:20:47] <tyzoid> didn't exit
[17:21:00] <deep42thought> hmm
[17:22:02] <deep42thought> yeah, for get-assignment, this is fine, I guess
[17:23:35] <tyzoid> It should exit, though
[17:23:43] <tyzoid> since this failure is triggered by not authenticating with buildmaster
[17:23:50] <tyzoid> i.e. if you don't have the ssh key
[17:23:51] <deep42thought> yes, I meant "exiting is fine"
[17:23:53] <tyzoid> ah
[17:23:56] <tyzoid> right
[17:24:06] <deep42thought> but for return-assignment, retrying is the way to go
[17:24:15] <deep42thought> because building packages is potentially expensive
[17:24:48] <tyzoid> True. At some point, it should die after a certain number of retries, so it can restart the build process
[17:25:16] <deep42thought> I don't think so
[17:25:41] <tyzoid> if it consistantly can't send the package back up to buildmaster, then something is wrong on the master or slave side
[17:25:44] <deep42thought> there is an exit code saying "I don't want your package"
[17:25:48] <tyzoid> first order of business would be to restart the slave
[17:25:57] <deep42thought> no
[17:26:00] <tyzoid> hence dying to allow auto-restart to kick in
[17:26:06] <deep42thought> first order of business is to repair the master :-)
[17:26:15] <deep42thought> e.g. the master goes offline
[17:26:25] <deep42thought> then the slave should try to resend again and again
[17:26:33] <tyzoid> from the point of view of the slave, it doesn't know if it's done something wrong or if the master is wrong
[17:26:41] <deep42thought> and only stop if the build master comes online again and says "I don't want your package"
[17:26:56] <deep42thought> if it's done something wrong, the master will tell it
[17:27:06] <tyzoid> Fair enough
[17:27:21] <deep42thought> e.g. "your package lacks signatures", "your split-package lacks parts", ...
[17:28:11] <deep42thought> and if the master does not recognize the error (and crashes with an unknown exit code), it's still the master's fault
[17:28:43] <deep42thought> only way I see, this can go wrong is if in the meantime something connection-related changes for the slave (e.g. you delete the ssh key)
[17:30:10] -!- bill-auger has quit [Read error: Connection reset by peer]
[17:30:47] -!- bill-auger has joined #archlinux32
[17:37:43] <buildmaster> twisted is broken (says nlopc46).
[17:42:28] -!- deep42thought has quit [Quit: Leaving.]
[18:21:42] <tyzoid> girls / deep42thought: the build slave I'm working on is at srv0.tyzoid.com:2209
[18:48:11] -!- bill-auger has quit [Quit: No Ping reply in 60 seconds.]
[18:48:54] -!- bill-auger has joined #archlinux32
[18:53:21] -!- bill-auger has quit [Ping timeout: 248 seconds]
[18:56:40] -!- bill-auger has joined #archlinux32
[19:00:58] -!- bill-auger_ has joined #archlinux32
[19:01:55] -!- bill-auger has quit [Ping timeout: 245 seconds]
[20:00:56] <buildmaster> openscenegraph is broken (says buildknecht).
[20:02:18] -!- bill-auger_ has quit [Ping timeout: 260 seconds]
[20:03:52] <abaumann> firefox: 274:03.78 ../../build/unix/gold/ld: fatal error: libxul.so: mmap: failed to allocate 1704008984 bytes for output file: Cannot allocate memory
[20:03:56] <abaumann> I'm aborting my slave..
[20:04:29] <abaumann> Disclaimer: I have 8 GB of RAM.
[20:08:06] -!- NoobAlice has joined #archlinux32
[20:19:54] <tyzoid> wat
[20:30:12] -!- NoobAlice has quit [Ping timeout: 245 seconds]
[20:33:06] -!- NoobAlice has joined #archlinux32
[20:40:28] -!- AndrevS has joined #archlinux32
[20:46:25] <tyzoid> abaumann: Did you see the chromium / ffmpeg issue on the forum?
[20:46:31] <abaumann> huh. no.
[20:46:58] <tyzoid> looks like the OP has it showing dependency on avcodec.so.57, whereas ffmpeg provides 58
[20:47:08] <abaumann> ah. we updated ffmpeg to 4 this morning (for mplayer)
[20:47:19] <tyzoid> yup, that'd do it.
[20:47:30] <tyzoid> audacity also needs a rebuild btw
[20:47:32] <abaumann> sad, it's only yet in testing.
[20:47:32] <tyzoid> if you haven't already
[20:47:59] <tyzoid> Not sure if it'll automatically catch the rebuild dep there, upstream missed that one
[20:48:14] <abaumann> let's see. It's a good test. :-)
[20:48:32] <tyzoid> it'll still work, it'll just fail to import mp3s/oggs
[20:48:36] <abaumann> anyway, I don't build browsers anymore on my slave, what a mess..
[20:49:10] <abaumann> I'm for an option '-np' in build-package: 'not-preferering'
[20:51:15] <abaumann> mmh. chromium is on the list of things to build. What worries me more, why is ffmpeg in stable..
[20:51:26] <abaumann> aeh.. ffmpeg 4.
[21:05:40] -!- deep42thought has joined #archlinux32
[21:05:41] <buildmaster> Hi deep42thought!
[21:06:26] <deep42thought> abaumann: as I mentioned on the forum: 8GB hdd (tmpfs) is not enough for firefox and thunderbird (on x86_64 crux, at least)
[21:07:55] <deep42thought> automatic rebuilds only happen with haskell-*, nothing else
[21:09:07] <tyzoid> wb deep42thought
[21:13:00] <abaumann> deep42thought: oups. missed that one. sorry.
[21:14:31] <deep42thought> nothing to be sorry about - I'm also surprised, that 8G are not enough :-)
[21:15:05] <abaumann> "give developers mighty machines and trouble you have with software.."
[21:15:14] <deep42thought> !grab abaumann
[21:15:14] <phrik> deep42thought: Tada!
[21:17:39] <tyzoid> deep42thought: only thing I'm not clear on, is where to put the packaging key
[21:17:53] <tyzoid> I assume non-password-protected in the build user's keyring?
[21:18:10] <deep42thought> exactly
[21:18:26] <tyzoid> Any chance we can allow password-protected keys?
[21:18:41] <deep42thought> how should that work?
[21:18:45] <tyzoid> i.e. when build-packages gets run, it asks for a password, and caches it in memory
[21:18:54] <tyzoid> that way, it's never stored unencrypted on disk
[21:19:08] <deep42thought> you can use gpg-agent
[21:19:24] <tyzoid> no idea how to set that up
[21:19:24] <deep42thought> (not tested)
[21:19:34] <tyzoid> never used it before
[21:19:34] <deep42thought> isn't that standard pretty much?
[21:19:52] <tyzoid> idk
[21:19:55] <deep42thought> you're being asked for your password every time?
[21:20:02] <tyzoid> oh, yes
[21:20:05] <tyzoid> each time it gets signed
[21:20:06] <tyzoid> yeah
[21:20:19] <tyzoid> but that requires I be present when something is going to be signed
[21:20:34] <deep42thought> https://wiki.archlinux.org
[21:20:34] <phrik> Title:GnuPG - ArchWiki (at wiki.archlinux.org)
[21:20:38] <tyzoid> might be a cool idea, actually
[21:22:10] <tyzoid> Oh, I think I get your comment - yes, by default it will cache
[21:22:23] <tyzoid> but it requires I sign something to get it cached
[21:22:54] <deep42thought> gpg -o - -e -r you /dev/null | gpg -d -o /dev/null -
[21:22:58] <deep42thought> or similar
[21:23:26] <deep42thought> umm, make "-e" to "-s" and you can spare the pipe
[21:23:34] <tyzoid> or I could just have a dummy file
[21:24:24] <deep42thought> just sign /dev/null
[21:25:27] <deep42thought> gpg -s -o /dev/null /dev/null
[21:25:53] <tyzoid> nice
[21:47:40] -!- isacdaavid has joined #archlinux32
[21:53:42] <buildmaster> coffeescript is broken (says tyzoid-srv0-bs0).
[21:54:15] <deep42thought> \o/
[21:54:19] <deep42thought> welcome to the team
[21:54:54] <tyzoid> lol
[21:55:08] <deep42thought> "mount: /var/lib/archbuild/staging-with-build-support-i686/root/proc: cannot mount proc read-only."
[21:55:10] <deep42thought> hmmm
[21:55:20] <deep42thought> maybe an emulation problem
[21:55:42] <tyzoid> journalctl -u build-slave.service, btw
[22:00:12] <tyzoid> May 24 19:54:00 srv0-arch32-build build-packages[1175]: 2018-05-24 19:54:00: building package "firefox" (revisions 63941b10f07c8281602d95597a99ca2e30e71ff1 83f1e7741b96dc02e7ea31e64d1cd580ef647b5b, repository extra, straw :clean_chroot:) ... failed.
[22:00:18] <tyzoid> deep42thought ^?
[22:00:28] <deep42thought> just status
[22:00:39] <tyzoid> Is that due to an issue with the builder? or is that normal?
[22:00:50] <tyzoid> (I want to see one successful build before I'm satisfied)
[22:01:00] <deep42thought> it's normal that it outputs status
[22:01:09] <deep42thought> but that the first failed so fast is not normal
[22:01:15] <deep42thought> look at the log
[22:01:16] <tyzoid> right
[22:01:30] <deep42thought> https://buildmaster.archlinux32.org
[22:03:15] <tyzoid> one sec, gonna try disabling that apparmour protection
[22:03:44] <deep42thought> bin/build-packages has a "-l" option which lets you try to do stuff locally
[22:03:46] <deep42thought> fyi
[22:04:08] <deep42thought> so "bin/build-packages -l coffeescript.4a95e63210eaf4837a9e0b5bea1c85e20aa866a3.7af66ce5d9207f3713d741e8cce402da73ba4fca.community" would redo the above build
[22:04:13] <deep42thought> locally
[22:09:56] <tyzoid> builder/bin/build-packages: line 415: sponge: command not found
[22:09:57] <tyzoid> gpg: skipped "0x15eebadc0de": Invalid user ID
[22:09:57] <tyzoid> gpg: signing failed: Invalid user ID
[22:10:01] <tyzoid> progress!
[22:10:06] <deep42thought> :-)
[22:10:13] <deep42thought> set the right id in conf/local.conf
[22:10:22] <buildmaster> aircrack-ng is broken (says buildknecht2).
[22:13:19] <tyzoid> deep42thought is sponge required?
[22:13:24] <deep42thought> yes
[22:13:30] <deep42thought> requirements are in README.md
[22:14:10] <tyzoid> !wtf sponge
[22:14:15] <phrik> tyzoid: community/moreutils
[22:15:43] -!- abaumann has quit [Quit: leaving]
[22:16:08] <tyzoid> deep42thought started it back up with systemd, heading off for now
[22:16:15] <deep42thought> cu later
[22:16:18] <tyzoid> If you want, feel free to stop it and test out some other packages
[22:16:27] <tyzoid> coffeescript encountered another error
[22:16:40] <tyzoid> 2018-05-24T20:12:56.build-log:Failed to attach 9442 to compat systemd cgroup /user.slice/user-1000.slice/session-7170.scope/payload: No such file or directory
[22:16:40] <tyzoid> 2018-05-24T20:12:56.build-log:Failed to attach 7910 to compat systemd cgroup /user.slice/user-1000.slice/session-7170.scope/supervisor: No such file or directory
[22:16:41] <tyzoid> 2018-05-24T20:12:56.build-log:Failed to chown() cgroup /sys/fs/cgroup/systemd/user.slice/user-1000.slice/session-7170.scope/payload: No such file or directory
[22:16:41] <tyzoid> 2018-05-24T20:12:56.build-log:==> ERROR: Aborting...
[22:16:42] <tyzoid> flock: 9: Bad file descriptor
[22:16:42] <tyzoid> Done.
[22:17:01] <tyzoid> so not sure whether it's the build-slave or the pkgbuild
[22:17:12] <tyzoid> regardless, see ya
[22:22:53] <eschwartz> this indicates your systemd was upgraded and weird subtle compatibility thingys ensued
[22:24:24] -!- deep42thought has quit [Quit: Leaving.]
[23:36:21] -!- AndrevS has quit [Quit: umount /dev/irc]
[23:56:03] -!- isacdaavid has quit [Quit: Leaving.]