#archlinux32 | Logs for 2023-06-08

Back
[00:05:15] -!- drathir_tor has joined #archlinux32
[05:08:08] <girls> you can, but if it's on one machine with not-horribly-much ram, I would recommend to only run one
[05:08:59] <girls> btw: no problem regarding getting side-tracked, I got distracted producing paperclips, too :)
[09:09:36] -!- JerryXiao has quit [Ping timeout: 268 seconds]
[09:11:18] -!- JerryXiao has joined #archlinux32
[10:00:16] -!- abaumann has joined #archlinux32
[10:00:16] <buildmaster> Hi abaumann!
[10:00:17] <buildmaster> !rq abaumann
[10:00:17] <phrik> buildmaster: <abaumann> wow: my first wayland experience of the blury kind (in libvirt).. there is a small black line moving from the top to the bottom of the screen.. an homage to old CRT monitors, I suppose. ;-)
[10:02:53] <abaumann> KitsuWhooa: I'm switching between a 6 slave on one machine doing -j1 and 1 slave on one machine going -j8. Hard to tell, what is faster..
[10:03:30] <abaumann> ..besides: a lot of packages just simple do a 'make -j' (or the rust/meson/whatever equivalent). So going with multiple build slaves risks to flood your machine.
[10:04:41] <abaumann> devtools32 will come, but the first prio is to get it to do what it did before in build-packages..
[10:05:03] <abaumann> ..I though about later adding an --arch32 flag, so that Arch32 specific patching can be done using pkgctl
[10:10:37] <abaumann> Aeh. This sedfu for staging.. ahem.. It should be enough to just have [staging] before all other [*-staging] and [testing] before all other [*-testing], right? ;-)
[10:15:49] <abaumann> copy_hostconf: urgh. overwrites the correctly set up servers..
[10:19:56] <abaumann> I fail to see how arch-nspawn ever worked in the past.
[10:21:06] <abaumann> it's picking mirrors (via extra) and them to mirrorlist and mirrorlist32 in the workdir
[10:24:35] <abaumann> ah. I hate git for auto-merging.
[10:24:40] <abaumann> it newer really works (TM)
[10:27:20] <abaumann> -host_mirrors=($(pacman-conf --repo extra Server 2> /dev/null | sed -r 's#(.*/)extra/os/.*#\1$repo/os/$arch#'))
[10:27:35] <abaumann> this is not respecting pac_conf passed via -C
[10:30:34] <abaumann> https://git.archlinux32.org
[10:30:36] <phrik> Title: devtools32 - Archlinux32 fork of devtools (at git.archlinux32.org)
[10:30:37] <abaumann> ok
[10:30:39] <abaumann> one by one :-)
[10:43:45] <abaumann> mmh, there is now core-staging-i486-build and extra-staging-i486-build, I think, I'll make a shin staging-i486-build for now to make the current build-package happy, can be adapted later..
[11:38:50] <abaumann> Cloning into bare repository '/home/slave1/builder/work/tmp.build-packages.efNxRq/binutils-gdb'...
[11:38:53] <abaumann> error: RPC failed; HTTP 408 curl 22 The requested URL returned error: 408
[11:38:56] <abaumann> fatal: expected 'packfile'
[11:38:58] <abaumann> ==> ERROR: Failure while downloading binutils-gdb git repo Aborting...
[11:39:01] <abaumann> mmh..
[11:40:04] <KitsuWhooa> abaumann: fair enough. Is there somewhere I can at least set MAKEFLAGS?
[11:40:14] <KitsuWhooa> most packages I've dealt with just run `make`
[11:40:19] <KitsuWhooa> not `make -j`
[11:40:32] <KitsuWhooa> so I'd at least like to have those build with MAKEFLAGS=-j20
[11:40:54] <KitsuWhooa> Also, who do I send the public pgp and ssh keys to?
[11:41:25] <abaumann> ah, those you send to me or deep42thought. We add them to the database on the buildmaster (so that it accepts your build slave).
[11:41:34] <abaumann> You also have to come up with a unique name :-)
[11:42:25] <abaumann> most packages respect the MAKEFLAGS (they get installed from devtools32 and can be adapted before installing them)
[11:42:54] <abaumann> there are some packages which have local patches in PKGBUILD which patch away make -j$(nproc) and similar craziness.
[11:43:17] <KitsuWhooa> I mean, that should be fine too
[11:43:22] <KitsuWhooa> the VM has 20 cores and 32GB RAM :p
[11:43:29] <abaumann> oh :-)
[11:43:32] <abaumann> sweet.
[11:44:14] <abaumann> mmh. huge packfiles in binutils?
[11:44:21] <abaumann> lemme play with git buffer sizes..
[11:46:55] <abaumann> deep42thought: you were building over tor because of restrictive gitlab settings upstream, riht?
[11:47:32] <abaumann> though, wait, this is binutils itself
[11:47:47] <abaumann> git clone --mirror https://sourceware.org /home/slave1/builder/work/tmp.build-packages.643CTy/binutils-gdb
[11:47:48] <phrik> Title: sourceware.org Git - binutils-gdb.git/summary (at sourceware.org)
[11:47:50] <abaumann> *sight*
[11:49:59] <abaumann> I really wish everybody would go back to tarball releasing..
[11:51:12] <abaumann> pkgctl repo clone --protocol=https --arch32 binutils-gdb
[11:51:18] <abaumann> asks me for a username?
[11:51:46] <abaumann> when it doesn't exist. :-)
[11:52:31] <abaumann> let's see what makepkg -do is doing..
[11:56:05] <abaumann> Cloning into bare repository '/home/abaumann/binutils/binutils-gdb'...
[11:56:05] <abaumann> error: RPC failed; HTTP 408 curl 22 The requested URL returned error: 408
[11:56:05] <abaumann> fatal: expected 'packfile'
[11:56:05] <abaumann> ==> ERROR: Failure while downloading binutils-gdb git repo Aborting...
[11:56:17] <abaumann> splendid. This is a binutils bug. I cannot download the sources..
[11:56:29] <abaumann> all my build slaves try to build the binutils package..
[11:56:44] <KitsuWhooa> ouch
[12:01:45] <abaumann> after all these years: RPC was a bad idea, webdav was and doing SCM over HTTPS as RPC is in no way superiour
[12:02:33] <abaumann> we had a "git sources dowload and cache" thingy on the buildmaster..
[12:02:44] <abaumann> ..but if the initial git clone is not working, what then..?
[12:03:01] <abaumann> I'm always puzzled upstream has not such issues..
[12:04:31] <abaumann> ui, why is git clone called with --mirror?
[12:05:10] <abaumann> for building packages this might be exagerated
[12:07:15] <KitsuWhooa> I'm at step 6 of https://bbs.archlinux32.org
[12:07:16] <phrik> Title: How to create a build slave? / Building / Arch Linux 32 Forum (at bbs.archlinux32.org)
[12:07:21] <KitsuWhooa> there's no mirrorlist32 package
[12:08:15] <KitsuWhooa> as for "create it by hand" just add that in /etc/pacman.d/mirrorlist32 ?
[12:08:38] <KitsuWhooa> oh there's already a file there
[12:08:58] <abaumann> yeah, on the most.
[12:09:01] <abaumann> *host
[12:09:11] <KitsuWhooa> ...host?
[12:09:17] <abaumann> where you build the packages on.
[12:09:19] <abaumann> :-)
[12:09:25] <abaumann> so, the slave ;-)
[12:09:25] <KitsuWhooa> yes
[12:09:35] <abaumann> matter of perspective :-)
[12:11:09] <abaumann> if ! git clone ${GITFLAGS:---mirror} "$url" "$dir"; then
[12:11:16] <abaumann> in makepkg
[12:11:37] <abaumann> This wa I get to:
[12:11:38] <abaumann> 2023-06-08 14:09:56: building package "binutils" (revisions a97d4a5818cdf0e5c7072dc475d0933d1e9e55fb ec4f7efbbe758d84e682ab6b9b881ccf92c3a82a, repository core, straw :with_build_support:clean_chroot:) for i686 ... failed.
[12:11:38] <KitsuWhooa> http://gr.mirror.archlinux32.org
[12:11:43] <KitsuWhooa> this returns a 404
[12:11:48] <abaumann> yes.
[12:12:00] <KitsuWhooa> is it supposed to?
[12:12:07] <KitsuWhooa> mirror_update_command='curl -v https://gr.mirror.archlinux32.org
[12:12:13] <abaumann> this is a refresh hook to make sure you have the local mirror in sync before building apackage.
[12:12:13] <KitsuWhooa> er
[12:12:15] <KitsuWhooa> *http
[12:12:23] <abaumann> it's just a trigger. :-)
[12:12:33] <KitsuWhooa> I'm not sure I understand
[12:12:36] <abaumann> you can ommit it, especially if you use mirror.archlinux32.org
[12:12:44] <KitsuWhooa> I picked gr.mirror
[12:12:49] <KitsuWhooa> because that's closest
[12:12:51] <abaumann> there is a rsync delay between mirror.archlinux32.org and gr.mirror
[12:12:52] <abaumann> ah.
[12:13:08] <KitsuWhooa> should I switch to mirror.arch[..]32.org?
[12:13:26] <KitsuWhooa> would've been the same to type mirror.archlinux32.org in terms of keystrokes
[12:13:32] <KitsuWhooa> but anyway, should i switch to that?
[12:13:33] <abaumann> then when say you build dependent packages, the buildmaster publishes on mirror.archlinux32.org, but you pick up files from the old staging from the gr.mirror.
[12:13:54] <KitsuWhooa> I see
[12:14:04] <abaumann> you can also use gr.mirror
[12:14:14] <abaumann> we had issues in the past especially around Haskell.
[12:14:20] <abaumann> but we no longer build it anyway..
[12:14:34] <KitsuWhooa> I'll switch to mirror.archlinux32.org to be safe then
[12:14:38] <abaumann> and most other software has a sane ABI and no checksums in .so files concept :->
[12:14:52] <abaumann> the worst which can happen is a build failure..
[12:15:05] <abaumann> ..so the buildmaster would do the build later anyway again.
[12:15:16] <abaumann> depends on your bandwidth.
[12:15:26] <abaumann> might be gr.mirror is better than mirror (which is in Germany)
[12:15:41] <KitsuWhooa> I think it'll be fine
[12:16:25] <abaumann> a, new pacman will have GITFLAGS in makekpkg, good. Problem solved then :-)
[12:16:28] <KitsuWhooa> Oh, lol. I just scrolled down a bit further and saw what that script is supposed to be
[12:16:32] <KitsuWhooa> sorry :p
[12:16:37] <abaumann> https://gitlab.archlinux.org
[12:16:38] <phrik> Title: scripts/libmakepkg/source/git.sh.in · master · Pacman / Pacman · GitLab (at gitlab.archlinux.org)
[12:16:48] <abaumann> easy :-)
[12:18:39] <KitsuWhooa> > builder/bin/build-package
[12:18:44] <KitsuWhooa> I assume that's meant to be build-packages
[12:22:09] <abaumann> "the candidate scores 100 points" :-)
[12:22:25] <KitsuWhooa> > The git repositories in work/repos/ are packages, community and packages32,
[12:22:28] <KitsuWhooa> I only have packages32
[12:22:50] <abaumann> yeah, I think, this is ok since the git move upstream
[12:22:56] <KitsuWhooa> alright
[12:23:19] <abaumann> we don't have complete git repo clones anymore (I think, but where is the data now? deep42thought might be the expert to ask here)
[12:23:37] <abaumann> Could be only a shallow git checkout during building in work/tmp.build-package.*
[13:01:21] <abaumann> deep42thought: there is Haskell building rechenknecht, this sounds suspicious. :-)
[13:03:56] <abaumann> + false
[13:03:56] <abaumann> + false
[13:03:56] <abaumann> + echo failure
[13:03:56] <abaumann> + printf ' failed.\n' failed.
[13:03:56] <abaumann> + false
[13:04:03] <abaumann> mmh. that's not so informative :-)
[13:04:49] <KitsuWhooa> hahaha
[13:05:16] <abaumann> + build_dir='/var/lib/archbuild/"${repository}-staging-with-build-support-pentium4/slave1'
[13:05:19] <abaumann> argh..
[13:05:27] <abaumann> my patch whoes.
[13:06:12] <abaumann> double-quotes inside single quotes, yack
[13:08:15] <abaumann> KitsuWhooa: on your build slave with a new devtools32 the build-packages script might want to start a staging-pentium4-build script which no longer exists, symlinking to core-staging-pentium4-build for now could be an idea.
[13:08:38] <abaumann> I'm patching the builder itself ATM
[13:09:06] <abaumann> "thuringiabuild1", oh. :-)
[13:09:09] <abaumann> a new one :-)
[13:09:28] <KitsuWhooa> abaumann: where would these files be?
[13:09:56] <abaumann> wherever you installed devtools32 with make install to /usr or /usr/local
[13:10:04] <abaumann> aeh 'bin'
[13:10:42] <KitsuWhooa> I didn't make install. I was told the package was up to date
[13:10:47] <KitsuWhooa> so I guess /usr/bin
[13:10:57] <abaumann> you have devtools32 installed?
[13:11:00] <KitsuWhooa> yes
[13:11:02] <abaumann> as a pacman package.
[13:11:05] <KitsuWhooa> yup
[13:11:07] <abaumann> ah. that's the older version.
[13:11:14] <abaumann> but then, it should work.
[13:11:30] <KitsuWhooa> 22:22:38 <KitsuWhooa> is devtools32 up to date or do I need to build it manually?
[13:11:32] <KitsuWhooa> 22:22:54 <girls> it's not more up-to-date in git
[13:11:34] <KitsuWhooa> 22:23:06 <girls> install the package, then run the sed, which I wrote above
[13:11:35] <abaumann> so it is not this bug (I have no devtools32 package and I installed git packages32 manually)
[13:11:51] <abaumann> that is correct
[13:12:00] <abaumann> I just checked in an hour ago into devtools32
[13:12:04] <KitsuWhooa> ah
[13:12:22] <KitsuWhooa> I do have /usr/bin/staging-pentium4-build
[13:12:36] <KitsuWhooa> I don't have core-staging-pentium4-build
[13:12:39] <abaumann> ah. I have a core-staging-with-build-support-pentium4-build running
[13:12:49] <KitsuWhooa> maybe symlink them the opposite way? :p
[13:12:52] <abaumann> but you have a staging-pentium4-build
[13:12:57] <abaumann> so this should be fine
[13:13:26] <abaumann> It's building on my build slave, but why it resorted to build-support is not so clear to me
[13:13:39] <KitsuWhooa> unfortunately I dont' know enough to help :p
[13:13:52] <abaumann> no worries :-)
[13:13:56] <abaumann> me neither ;-)
[13:14:00] <KitsuWhooa> :D
[13:14:19] <abaumann> but, now, when I get a binutils built, you might get different assignments which are probably easier to build
[13:14:55] <KitsuWhooa> we'll see
[13:16:37] <KitsuWhooa> you're building pentium4 not i686
[13:16:43] <KitsuWhooa> so maybe I won't get anything easier
[13:16:44] <KitsuWhooa> :p
[13:16:58] <abaumann> yup
[13:18:24] <abaumann> btw: you can live-watch the build with tail -f tmp.build-packages.*/*build-log
[13:18:27] <abaumann> (in work)
[13:18:52] <KitsuWhooa> ah nice
[13:21:49] <abaumann> don't press ctrl-c in the wrong place (as I did) *grmpf*
[13:22:51] <KitsuWhooa> I'll try to keep it in mind
[13:23:02] <KitsuWhooa> but yeah, I'm still stuck looping on i686/binutils
[13:23:41] <KitsuWhooa> https://archlinux32.org
[13:23:50] <KitsuWhooa> what's the "unknown" failure?
[13:24:00] <KitsuWhooa> sudo: sorry, you are not allowed to set the following environment variables: MAKEFLAGS, PKGEXT
[13:24:19] <KitsuWhooa> Oh I just got error: RPC failed; HTTP 408 curl 22 The requested URL returned error: 408
[13:24:21] <KitsuWhooa> :p
[13:24:31] <abaumann> yep. remove the --mirror in makepkg
[13:24:37] <abaumann> in the git cone
[13:24:39] <abaumann> *clone
[13:25:25] <abaumann> aeh. sorry. it's in /usr/share/makepkg/source/git.sh
[13:25:45] <abaumann> will be a GITFLAGS which can override the --mirror default
[13:25:50] <abaumann> *there will
[13:26:28] <KitsuWhooa> Ooooh
[13:26:46] <KitsuWhooa> let's try it again
[13:27:01] <abaumann> I can imagine they left the --mirror there so that you can cherry-pick git patches from other branches and such.
[13:27:20] <KitsuWhooa> I assume it's going to get nuked next time I pacman -Syu and there's a new version
[13:27:21] <abaumann> but binutils is quite big, so the git repo server might choke to deliver a full mirror
[13:27:34] <abaumann> positive :-)
[13:27:53] <abaumann> only really big packages are affected anyway
[13:28:00] <abaumann> I also suspect, it's binutils only.
[13:28:39] <KitsuWhooa> it just failed again...
[13:29:48] <KitsuWhooa> how often are the build logs updated?
[13:30:04] <abaumann> mmh. after a complete build, I think
[13:30:20] <KitsuWhooa> I refreshed the page above but I still see the same failure log
[13:30:42] <abaumann> archlinux32.org is a mysql mirror of the buildmaster
[13:30:47] <KitsuWhooa> oh
[13:30:49] <abaumann> but it should be fairly current
[13:30:55] <abaumann> there is a mysql replication there
[13:31:07] <KitsuWhooa> https://buildmaster.archlinux32.org then?
[13:31:08] <phrik> Title: Index of /build-logs/error (at buildmaster.archlinux32.org)
[13:31:23] <abaumann> should be the master, yes
[13:31:34] <KitsuWhooa> what's the timezone? UTC?
[13:31:36] <abaumann> just without nice interface
[13:32:23] <abaumann> oeh, CEST
[13:32:29] <KitsuWhooa> ah
[13:32:53] <abaumann> gr is EEST?
[13:33:01] <abaumann> so one our ahead?
[13:33:01] <KitsuWhooa> yup
[13:33:03] <KitsuWhooa> one hour ahead
[13:33:16] <KitsuWhooa> Could this be the reason it's failing? > sudo: sorry, you are not allowed to set the following environment variables: MAKEFLAGS, PKGEXT
[13:33:19] <KitsuWhooa> I can't find any other logs
[13:34:10] <KitsuWhooa> I assume that's a warning...
[13:34:36] <abaumann> it might be easier to set them in /usr/share/devtools
[13:34:53] <abaumann> those makepkg-pentium4.conf etc. files are the templates when new build chroots are created with mkchroot
[13:35:10] <KitsuWhooa> maybe I can just tell sudo to let them through for now
[13:35:13] <abaumann> so, if you change something, better delete the chroots in /var/lib/archbuild
[13:35:23] <abaumann> this is also an option of course
[13:35:27] <KitsuWhooa> I do not have /var/lib/archbuild
[13:35:44] <KitsuWhooa> I assumed there was no reason to set that up manually if I didn't want more than one
[13:35:45] <KitsuWhooa> right?
[13:36:06] <abaumann> it should appear when it creates the first chroot, yes
[13:36:22] <KitsuWhooa> ls: cannot access '/var/lib/archbuild': No such file or directory
[13:37:46] <abaumann> anyway.. binutils fails happily in some gold linker tests. :-)
[13:38:43] <abaumann> mmh. but it built a binutils. currently uploading..
[13:39:05] <KitsuWhooa> Looks like allowing the env vars through sudo is getting me further
[13:39:10] <abaumann> I hope, we are not using the gold linker in too many places (it runs out of virtual memory anyway)
[13:39:37] <KitsuWhooa> it's running configure!
[13:41:25] <abaumann> good :-)
[13:41:47] <abaumann> gzip: stdin: unexpected end of file
[13:41:47] <abaumann> tar: Child returned status 1
[13:41:47] <abaumann> tar: Error is not recoverable: exiting now
[13:41:47] <abaumann> ==> ERROR: pkgrel is not allowed to be empty.
[13:41:47] <abaumann> ==> ERROR: pkgver is not allowed to be empty.
[13:41:49] <abaumann> ==> ERROR: pkgname is not allowed to be empty.
[13:41:52] <abaumann> on gcc.
[13:41:56] <abaumann> oh, this is bad.
[13:42:17] <abaumann> https://archlinux32.org
[13:42:18] <phrik> Title: Arch Linux 32 - List of Build Slaves (at archlinux32.org)
[13:42:25] <abaumann> yeah, binutils i686 is coming along nicely :-)
[13:44:11] <abaumann> so, binutils ended up in core-staging and no longer in staging.
[13:44:23] <abaumann> -rw-r--r-- 1 http http 7960177 Jun 8 15:40 binutils-2.40-6.0-pentium4.pkg.tar.zst
[13:44:26] <abaumann> good.
[13:54:53] <KitsuWhooa> abaumann: 16:54:31 <buildmaster> Your buildslave "build-tasossah" uploaded a package with a not fully-trusted signature:
[13:55:29] <KitsuWhooa> I think I might've made an oopsie
[13:56:00] <KitsuWhooa> or not
[13:56:04] <KitsuWhooa> I think you didn't trust my key
[13:56:10] <abaumann> ah, this just means that the package signing check doesn't trust you key.
[13:56:12] <abaumann> exactly :-)
[13:56:18] <KitsuWhooa> and now I'm rebuilding binutils!
[13:56:20] <KitsuWhooa> :p
[13:56:27] <abaumann> oupsi
[13:56:37] <abaumann> wasted electrons.. once more ;-)
[13:57:09] <abaumann> aeh.. lemme see. I have to import your signing key also somewhere..
[14:05:44] <abaumann> oh. and your public key has to be added to the archlinux32-keyring, otherwise installers might be extremly happy.
[14:06:08] <KitsuWhooa> yeah I was thinkin about that
[14:06:11] <KitsuWhooa> *thinking
[14:06:38] <abaumann> the internet is full of my email address and broken gpg complaints. :-)
[14:11:34] <KitsuWhooa> another round of building binutils :p
[14:11:48] <KitsuWhooa> Maybe I should stop the builder until you've got it sorted out
[14:12:24] <abaumann> or you let it run to warm up the electrons.. ;-)
[14:12:36] <abaumann> I still have trouble to remember all I have to do..
[14:12:46] <abaumann> ..so I should maybe also start a documentaiton.
[14:12:48] <KitsuWhooa> if it wasn't my desktop, maybe I wouldn't mind
[14:12:50] <KitsuWhooa> :p
[14:13:10] <abaumann> as I said, you can stop any time
[14:13:25] <abaumann> I have to upload the pupkey to our keyserver
[14:13:32] <KitsuWhooa> I just did. Let me know when to try again
[14:13:43] <abaumann> then make a new keyring-package with the key in it. update the buildmaster. that should be it
[14:13:47] <abaumann> ok.
[14:26:03] <abaumann> your key lives now here: https://archlinux32.org
[14:26:17] <abaumann> but we have to go through the hoops of trusting it now, I'm afraid
[14:26:17] <KitsuWhooa> nice
[14:26:54] <KitsuWhooa> surely it's not that ard
[14:26:56] <KitsuWhooa> *hard
[14:26:57] <abaumann> archlinux32-keyring, update-keys is not trusting your key, but that's prerequisite
[14:27:07] <abaumann> for me - always :-)
[14:27:09] <KitsuWhooa> :D
[14:27:14] <abaumann> I don't have a PGP/crypto-mind
[14:29:39] <abaumann> https://git.archlinux32.org
[14:29:41] <phrik> Title: archlinux32-keyring - Archlinux32 pacman keyring (at git.archlinux32.org)
[14:30:01] <abaumann> what puzzles me is that update-keys is updating (and revoking) a little bit too much for my taste
[14:35:30] <abaumann> "If you do not have one yet, generate it now, get it signed by at least three devs and get it introduced in archlinux32-keyring (the last two points are basically the same)."
[14:35:48] <abaumann> the at least three devs is not really possible ATM
[14:35:52] <KitsuWhooa> I was about to say
[14:43:28] <abaumann> we have to do something like this fist:
[14:43:30] <abaumann> https://gist.github.com
[14:44:27] <abaumann> I have to admit: I so much hate GPGs way of dealing with things.. it's always almost impossible for me to use it.
[14:44:55] <abaumann> the three signers is a requirement by gpg?
[14:45:14] <abaumann> so, I signed and uploaded your key. But I'm afraid deep42thought must also do this..
[14:45:20] <abaumann> key is not fully trusted: 80EC18799E8BCD375C6E64ABE4D41569196B1160
[14:45:32] <abaumann> so let's postpone the rest.. as I have to leave now..
[14:45:36] <KitsuWhooa> sure
[14:46:10] <KitsuWhooa> when/if you do figure it out, let me know what else I need to do
[14:46:16] <abaumann> ok.
[14:46:48] <abaumann> cu.
[14:46:52] <KitsuWhooa> see ya
[14:46:54] -!- abaumann has quit [Quit: leaving]
[14:54:38] -!- City-busz has quit [Quit: http://quassel-irc.org - Csevegjen kényelmesen. Bárhol.]
[15:35:05] -!- T`aZ has quit [Ping timeout: 265 seconds]
[15:36:52] -!- T`aZ has joined #archlinux32
[16:13:47] -!- T`aZ has quit [Ping timeout: 265 seconds]
[16:15:22] -!- T`aZ has joined #archlinux32
[16:27:33] <girls> yeah, the build master checks, that the key is enough-trusted by our (current) keyring
[16:27:43] <girls> so we don't accidentally upload uninstallable packages
[17:06:30] <KitsuWhooa> so, what now? :p
[17:07:11] <girls> I need to figure out, how to start scdaemon on my box
[17:07:14] <girls> and sign your key
[17:07:19] <girls> abaumann needs to do the same
[17:07:22] <girls> and we need a third
[17:07:36] <girls> I'll ask citybusz and polichronucci
[17:10:55] <KitsuWhooa> I see
[17:12:07] -!- T`aZ has quit [Ping timeout: 240 seconds]
[17:14:13] -!- T`aZ has joined #archlinux32
[17:16:10] <girls> damn, how hard can it be to get scdaemon running :(
[17:23:35] <girls> aha! got it :)
[17:23:51] <girls> forgot to start pcscd, too
[17:39:57] <girls> abaumann: no wonder, there were so many changes in the keyring - I forgot to push my last changes
[17:40:06] <girls> $ ./update-keys
[17:40:06] <girls> gpg: Generating Arch Linux 32 Keyring keychain master key...
[17:40:06] <girls> gpg: Done
[17:40:06] <girls> key is not fully trusted: 80EC18799E8BCD375C6E64ABE4D41569196B1160 KitsuWhooa
[17:40:22] <girls> ok, now we need to wait for citybusz or polichronucci's signatures :)
[17:40:48] <KitsuWhooa> haven't seen either in a while :p
[17:41:17] <girls> no trouble, they usually react reasonably fast, but don't hang around in irc all day ;)
[17:42:34] <KitsuWhooa> I'm curious though. How would one set this up if it was just one person?
[17:42:44] <KitsuWhooa> Just have random people sign the key?
[17:42:48] <girls> then one person would hold three keys
[17:42:51] <KitsuWhooa> ahahaha
[17:42:53] <KitsuWhooa> I see
[17:42:59] <girls> it's just abusing gpg
[17:43:11] <girls> crux made the sane decision and went for signify
[17:43:30] <girls> which is plain "here is a pubkey", "here is a signature" - does it match?
[18:00:05] <KitsuWhooa> mhm
[18:49:20] -!- drathir_tor has quit [Ping timeout: 240 seconds]
[19:05:35] -!- drathir_tor has joined #archlinux32
[19:08:39] -!- drathir_tor has quit [Remote host closed the connection]
[19:20:38] -!- drathir_tor has joined #archlinux32
[19:42:08] -!- drathir_tor has quit [Ping timeout: 240 seconds]
[19:49:35] -!- drathir_tor has joined #archlinux32
[20:28:56] -!- drathir_tor has quit [Ping timeout: 240 seconds]
[20:31:23] -!- drathir_tor has joined #archlinux32
[20:43:44] -!- drathir_tor has quit [Ping timeout: 240 seconds]
[20:59:43] -!- GNUtoo has quit [Remote host closed the connection]
[21:05:28] -!- drathir_tor has joined #archlinux32
[22:11:44] -!- drathir_tor has quit [Ping timeout: 240 seconds]
[22:36:18] <bill-auger> presumably, the intuition there, is that a lone package maintainer is not sufficient to manage an entire distro
[22:44:56] <bill-auger> also, there is no "web of trust", if no one has ever met that a lone package maintainer
[22:53:54] -!- drathir_tor has joined #archlinux32
[23:18:08] -!- drathir_tor has quit [Ping timeout: 240 seconds]
[23:20:09] -!- drathir_tor has joined #archlinux32
[23:25:33] -!- drathir_tor has quit [Remote host closed the connection]
[23:25:59] -!- drathir_tor has joined #archlinux32