#archlinux32 | Logs for 2019-06-12

[00:32:15] -!- finsternis has quit [Excess Flood]
[00:32:46] -!- finsternis has joined #archlinux32
[00:36:26] -!- isacdaavid has joined #archlinux32
[00:41:43] -!- ofara_ has quit [Ping timeout: 245 seconds]
[00:42:47] -!- ofara_ has joined #archlinux32
[01:02:23] -!- isacdaavid has quit [Quit: Leaving.]
[01:34:55] -!- eschwartz has joined #archlinux32
[02:00:23] -!- davor_ has joined #archlinux32
[02:01:26] -!- davor has quit [Ping timeout: 272 seconds]
[02:01:26] davor_ is now known as davor
[02:18:13] -!- ubuntuxp has joined #archlinux32
[02:40:04] -!- eschwartz has quit [Ping timeout: 272 seconds]
[02:52:54] -!- thePiGrepper has joined #archlinux32
[03:07:52] -!- eschwartz has joined #archlinux32
[05:49:42] -!- ubuntuxp has quit [Ping timeout: 252 seconds]
[06:17:22] -!- oaken-source has quit [Ping timeout: 244 seconds]
[06:17:38] -!- oaken-source has joined #archlinux32
[08:34:09] -!- quequotion_ has joined #archlinux32
[08:34:54] -!- abaumann has joined #archlinux32
[08:34:54] <buildmaster> Hi abaumann!
[08:34:54] <buildmaster> !rq abaumann
[08:34:55] <phrik> buildmaster: <abaumann> doesn't ## uncomment the #? ;-)
[08:42:51] <abaumann> girls: welcome to the modern world of connectivity. ;-)
[08:48:48] -!- deep42thought has joined #archlinux32
[08:48:48] <buildmaster> Hi deep42thought!
[08:48:48] <buildmaster> !rq deep42thought
[08:48:49] <phrik> buildmaster: <deep42thought> one day, some genius will prove, that all programming languages are isomorphic and there is actually only _one_ compiler
[08:50:30] <abaumann> hi deep42thought
[08:50:49] <deep42thought> Hi abaumann!
[09:04:59] <deep42thought> I should add this as a banner for all those who try to connect via ipv6 https://tnx.nl (sadly, this also applies to the master mirror)
[09:07:12] <abaumann> I have the same problem with my ISP, they don't provide IPv6
[09:07:39] <abaumann> I like the warning sign :-)
[09:07:47] <deep42thought> yeah :D
[09:11:29] -!- thePiGrepper has quit [Ping timeout: 272 seconds]
[09:28:53] <deep42thought> hmm, I can't think of a way to display this sign to users who only have ipv6 without breaking connectivity for dual-stack users :-/
[10:07:20] <T`aZ> hmm, im somehow using pentium4 , but when building from aur it builds foo-i686 and then at install complains that the arch i686 is not compatible
[10:07:37] <deep42thought> have you merged your /etc/makepkg.conf?
[10:08:06] <T`aZ> of coruse not!, let me take a look
[10:10:31] <abaumann> 1500->1178 packages to be build, in no time..
[10:10:44] <abaumann> ..before basically a flat-line..
[10:10:45] <deep42thought> I fixed a problem with get-package-updates
[10:10:49] <abaumann> ah. :-)
[10:11:00] <T`aZ> I'm only seeing CARCH="pentium4" change
[10:11:05] <deep42thought> nvidia-340xx* was dropped upstream and caused get-package-updates to error out
[10:11:13] <T`aZ> palemoon fails to build with that
[10:11:14] <deep42thought> T`aZ: that's the relevant one
[10:11:24] <deep42thought> so you have CARCH=pentium4?
[10:11:48] <T`aZ> i did not, i tried, but palemoon-bin failed
[10:12:04] <T`aZ> so i switched to i686 in pacman.conf just for it
[10:12:11] <deep42thought> there's probably a $CARCH switch somewhere in the PKGBUILD that needs to be extended to pentium4
[10:12:31] <deep42thought> e.g. a binary which gets downloaded (source_i686)
[10:12:37] <abaumann> palemoon-bin is merely downloding some binary blob and installs it.
[10:12:58] <deep42thought> we might want to add fallback variables for arch=()=
[10:12:59] <deep42thought> ?
[10:13:02] <T`aZ> yes, the download failed, i haven't looked at why
[10:13:18] <deep42thought> e.g. if it has 'arch=(i686)', then makepkg should happily build for "pentium4", too
[10:13:36] <deep42thought> though it would make things overly complex
[10:13:37] <deep42thought> :-/
[10:13:44] <abaumann> source_i686=("https://linux.palemoon.org/datastore/release/palemoon-$pkgver.linux-i686.tar.bz2"{,.sig})
[10:13:49] <abaumann> needs a source_pentium4
[10:14:15] <abaumann> otherwise you don't download anything and cp and install and empty directory :-)
[10:14:24] <abaumann> sha256sums_pentium4 dito
[10:19:32] <abaumann> palemoon-bin: built and works, requires SSE2 (at least this binary version)
[10:19:49] <deep42thought> so it should be "pentium4" all the way long, not "i686"
[10:20:12] <abaumann> but in the download link: linux-i686
[10:20:29] <deep42thought> yes
[10:21:00] <deep42thought> like someone once said: we should have i686 require sse2 and introduce something below, that does not require it ...
[10:21:11] <deep42thought> but now it's too late for that, I think
[10:21:39] <abaumann> yeah
[10:22:27] <abaumann> http://archlinux32.andreasbaumann.cc
[10:22:32] <abaumann> http://archlinux32.andreasbaumann.cc
[10:22:39] <abaumann> http://archlinux32.andreasbaumann.cc
[10:24:59] <deep42thought> for some reason, I do not receive status emails about failing systemd jobs (anymore?) ... this makes my latency rather high
[10:25:15] <abaumann> from the buildmaster?
[10:25:18] <deep42thought> yes
[10:25:22] <abaumann> mmh
[10:25:44] <deep42thought> I'm afraid, it is the last stage that fails: actually committing the emails to some mail server
[10:26:00] <abaumann> which mailer are you using?
[10:26:05] <deep42thought> msmtpd
[10:26:13] <deep42thought> msmtp
[10:26:32] <deep42thought> but there are some scripts working before that to collect emails into digests
[10:26:39] <deep42thought> and to sign/encrypt them
[10:26:40] <abaumann> digest-mailer
[10:26:43] <deep42thought> right
[10:27:20] <deep42thought> I'll have to dig ...
[10:27:58] <abaumann> /usr/bin/digest-mailer: line 164: sendmail: command not found
[10:28:20] <deep42thought> indeeeeeed
[10:28:30] <abaumann> shouldn't that be part of postfix?
[10:28:52] <abaumann> this looks pretty much like a cleanup of... hum.. maybe me
[10:28:56] <deep42thought> you have postfix installed?
[10:28:59] <abaumann> now
[10:29:16] <deep42thought> isn't that a little much for *just* sending emails?
[10:29:23] <abaumann> yeah.
[10:29:39] <abaumann> but otoh sooner or later I usually install a postfix anyway for other reasons..
[10:29:42] -!- T`aZ has quit [Remote host closed the connection]
[10:29:59] <deep42thought> ok
[10:30:05] -!- phrik has quit [Read error: Connection reset by peer]
[10:32:00] <deep42thought> I see hundrets of haskell-* packages rolling in - this will be fun :-)
[10:32:01] <abaumann> I see tons of connect to mail.<yourdomain> connection refused
[10:32:23] <abaumann> so, your mail server has to accept mails from the buildmaster, I think
[10:34:29] -!- quequotion_ has quit [Quit: quequotion_]
[10:34:33] <deep42thought> it should ...
[10:38:58] -!- phrik has joined #archlinux32
[10:38:58] <buildmaster> Hi phrik!
[10:38:58] <buildmaster> !rq phrik
[10:42:38] <phrik> buildmaster: <phrik> ayy
[10:43:56] <abaumann> mmh. coonection refused to my mailserver too.
[10:44:09] <deep42thought> maybe a "feature" of your provider?
[10:44:16] <deep42thought> some tend to block port 25 connections
[10:44:19] <abaumann> if, it's time to change the provider.
[10:44:23] <deep42thought> :-D
[10:45:49] <abaumann> ok. so we have to go over 587
[10:46:00] <deep42thought> I have only 465 open
[10:46:10] <deep42thought> (smtps)
[11:08:28] -!- davor_ has joined #archlinux32
[11:10:57] <buildmaster> pentium4/virtualbox-modules-arch is broken (says eurobuild3).
[11:11:37] -!- davor has quit [Ping timeout: 264 seconds]
[11:11:38] davor_ is now known as davor
[11:17:58] -!- abaumann has quit [Ping timeout: 246 seconds]
[11:42:57] -!- AndrevS has joined #archlinux32
[11:43:12] <buildmaster> pentium4/adrdox is broken (says rechenknecht).
[12:40:48] -!- abaumann has joined #archlinux32
[12:40:49] <buildmaster> Hi abaumann!
[12:40:49] <buildmaster> !rq abaumann
[12:40:49] <phrik> buildmaster: <abaumann> well, we found out that those disks _ARE_ actually slow disks then :-)
[12:40:56] <buildmaster> pentium4/perl-class-load-xs are broken (says rechenknecht).
[12:40:57] <abaumann> mmh. 3471 packages to be built..
[12:43:16] <abaumann> I patched some of webkit2gtk, but I'm afraid, it's just enough to keep other packages on top building. For actually working there are several micro-optimizations for SSE2 which have to be either translated to plain C or have to be done otherwise.
[13:01:37] <abaumann> I'm rerouting the 587 port of the buildmaster/postfix to my own mailserver, which then forwards the mails also to your mail servers
[13:01:48] <abaumann> you should get status emails now from the buildmaster
[13:12:14] <buildmaster> pentium4/perl-inline-java is broken (says rechenknecht).
[13:14:24] <deep42thought> abaumann: indeed, I see (the) emails from the buildmaster now :-)
[13:14:43] <abaumann> no too many, I hope. ;-)
[13:15:04] <deep42thought> only two: one test email I sent, and another, empty one about mariadb :-/
[13:15:53] <abaumann> ah. systemd-email.
[13:16:04] <abaumann> it does something I don't understand: Failed to connect to bus: No such file or directory
[13:16:09] <abaumann> as user master
[13:16:29] <deep42thought> systemctl is aliased to 'systemctl --user'
[13:16:39] <deep42thought> not sure if this is relevant for your issue
[13:17:09] <abaumann> ah. it works when I'm unaliasing the alis.
[13:17:22] <deep42thought> you're using 'sudo systemctl ...'?
[13:17:42] <deep42thought> the build master scripts run in the user context of user master
[13:17:44] <abaumann> no, I just tried to run the script
[13:31:16] <buildmaster> pentium4/perl-term-readline-gnu is broken (says rechenknecht).
[13:47:11] <deep42thought> hmm, that's interesting: a new perl (5.30) is rolling into testing, but I strongly doubt that all perl-* packages have already been recompiled
[13:48:05] -!- ofara_ has quit [Quit: ofara_]
[13:48:26] -!- ofara_ has joined #archlinux32
[13:52:41] <abaumann> yeah. tons of perl things are broken, for example libguestfs
[13:52:50] <abaumann> perl-cache-cache
[13:53:04] <deep42thought> sure, most probably every single perl package :-)
[13:54:08] <abaumann> I have to check, some things could also have come from the AUR
[13:54:20] <abaumann> libguestfs is from the AUR
[13:54:26] <abaumann> perl-sys-virt seems fine
[13:55:15] <abaumann> hivex is AUR too
[13:55:35] <abaumann> some libvirt stuff on my buildslave are broken.. fixing..
[13:55:48] <deep42thought> pidgin and libpurple have /funny/ dependencies in the database :-/
[13:56:47] <deep42thought> they depend on 'perl>=i686-linux-thread-multi.' and 'perl<i686-linux-thread-multi.1' (names of the install targets) and the version information is empty
[13:58:30] <abaumann> mmh. some shell trickery in the depends section?
[13:58:43] <deep42thought> probably
[14:04:20] <deep42thought> oh, this is my own sed-fu-trickery
[14:07:57] <deep42thought> aha! pidgin actually comes with the directory usr/lib/perl5/i686-linux-thread-multi/
[14:11:37] <abaumann> oh.
[14:11:48] <deep42thought> the directory is empty
[14:12:02] <deep42thought> I just made the check what we accept as a perl "version" more strict
[14:12:11] <deep42thought> it now must consist of [0-9.] only
[14:22:17] -!- ahuillet has joined #archlinux32
[14:22:28] <ahuillet> mmh
[14:22:32] <ahuillet> warning: glibc: local (2.29-1.26) is newer than core (2.29-1.3)
[14:22:32] <ahuillet> and more
[14:22:40] <ahuillet> some mirror outdated?
[14:22:50] <deep42thought> which are you currently on?
[14:23:32] <ahuillet> archlinux32.agoctrl.org on which I ran syncrepo.sh 10 seconds before joining the channel
[14:23:42] <ahuillet> source_url='rsync://mirror.archlinux32.org/archlinux32'
[14:23:50] <deep42thought> that one should be up-to-date
[14:24:47] <deep42thought> you switched from i686 to pentium4?
[14:24:49] <deep42thought> possibly?
[14:25:00] <ahuillet> I didn't do anything on purpose, but maybe that is happening
[14:25:07] <deep42thought> pacman-conf Architecture
[14:25:08] <ahuillet> eeePC 701, not sure if it qualifies as pentium4 even
[14:25:13] <ahuillet> pentium4.
[14:25:20] <deep42thought> then it does :-)
[14:25:24] <deep42thought> https://news.archlinux32.org
[14:25:25] <phrik> Title: Archlinux 32 News (at news.archlinux32.org)
[14:26:32] <ahuillet> oh yes thank you, sorry I didn't pay attention that it wasn't fully automated
[14:26:43] <ahuillet> full package reinstall on a 4GB / is gonna be fun.
[14:27:04] <deep42thought> you have the cache on that 4GB, too?
[14:27:18] <deep42thought> maybe it's a good idea to put that on an external drive for the moment
[14:34:11] <abaumann> not symlinking it.. ;-)
[14:35:02] <deep42thought> symlinking the cache should not be a problem, or?
[14:35:32] <deep42thought> ah, no it's owned by pacman, so it _is_ a problem :-)
[14:35:36] <abaumann> *abaumann caughs nevously and points at recent disaster not configuring 'CacheDir' in /etc/pacman.conf
[14:35:44] <abaumann> *nervously
[14:37:32] <ahuillet> cache is elsewhere thankfully
[14:45:07] <ahuillet> even with a bunch of NoExtract rules the operation seems to require about 500MB of free space just for applying the "upgrades". that's going to be fun. 4GB is really kind of small these days.
[14:45:39] <abaumann> yeah. I'm overlaying half of my installation on my eeepc from the 4gb to a external 32gb sd card.
[14:46:07] <ahuillet> I have / on internal drive and external SD for data storage (/home) but maybe that needs to change.
[14:47:40] <abaumann> /dev/sda1 3.7G 3.4G 142M 97% /
[14:47:43] <abaumann> *urgh*
[14:47:45] <buildmaster> pentium4/tokei is broken (says rechenknecht).
[14:51:28] -!- quequotion has joined #archlinux32
[14:52:26] <elibrokeit> ahuillet: I'm not positive but I suspect that pacman's "checking for free space" step does not know how to calculate how much space from a package is negated by a NoExtract rule ;)
[14:53:06] <elibrokeit> HOWEVER, there is a "solution"
[14:53:18] <abaumann> disable the check? ;-)
[14:53:23] <elibrokeit> You can comment out "CheckSpace" in the default pacman.conf, to stop pacman from checking at all
[14:53:26] <elibrokeit> yep
[14:53:47] <deep42thought> what does noextract in context of installing a package mean?
[14:53:47] <elibrokeit> that will mean you're in charge of accounting for your own disk space, so Here Be Dragons.
[14:54:19] <elibrokeit> deep42thought: it means even though the package installs the file, pacman on a specific system will simply skip it when extracting files
[14:54:40] <elibrokeit> it also means an upgrade will delete the file, if it exists
[14:54:45] <deep42thought> ah, so I can skip files from installation
[14:54:52] <deep42thought> thanks
[15:04:31] -!- deep42thought has quit [Quit: Leaving.]
[15:09:35] -!- eschwartz has quit [Ping timeout: 258 seconds]
[15:19:33] <buildmaster> i686/haskell-contravariant-extras are broken (says eurobuild1).
[15:22:36] -!- abaumann has quit [Quit: leaving]
[15:33:39] <ahuillet> elibrokeit: I don't want it to fail because out of disk space though
[15:35:25] <ahuillet> so I'll do the reinstall in N chunks instead
[15:38:40] -!- thePiGrepper has joined #archlinux32
[15:51:06] <elibrokeit> ahuillet: if you have enough space... :)
[15:51:35] <ahuillet> chunking should work! I hope!
[15:55:56] <elibrokeit> But so should disabling the check
[15:55:59] <elibrokeit> :p
[15:56:15] <elibrokeit> Actually we do exactly that when building in chroots
[15:56:35] <elibrokeit> What is the worst that happens if the chroot breaks anyway!
[15:57:03] <ahuillet> I'm not completely sure that it would work honestly, at least I'd rather not get to a full FS
[15:57:10] <ahuillet> that happened recently and it's hell to recover from
[16:09:55] <elibrokeit> True
[16:10:35] <elibrokeit> Which packages are adding up to 500MB of removrd NoExtract files?
[16:11:14] <ahuillet> I have no idea. if you have quick way to find out I'll tell you
[16:11:54] <ahuillet> https://pastebin.aquilenet.fr
[16:11:55] <phrik> Title: PrivateBin (at pastebin.aquilenet.fr)
[16:12:15] <ahuillet> and I think I probably could add all of /usr/doc for good measure
[16:14:30] <elibrokeit> Your pastebin doesn't work with my Android browser.
[16:14:40] <ahuillet> sorry about that
[16:14:46] <elibrokeit> It's asking for a decryption key :/
[16:14:46] <ahuillet> NoExtract = usr/share/help/* !usr/share/help/en* !usr/share/help/fr* usr/share/locale/* !usr/share/locale/en/* !usr/share/locale/en_GB/* !usr/share/locale/en_US/* !usr/share/locale/fr* !usr/share/locale/kf5_all_languages !usr/share/locale/locale.alias usr/share/man/* !usr/share/man/man* usr/share/doc/HTML/* usr/share/X11/locale/* !usr/share/X11/locale/compose.dir !usr/share/X11/locale/locale* !usr/share/X11/locale/en_
[16:14:46] <ahuillet> US.UTF-8/* usr/lib/python2.7/test/* usr/lib/python3*/test/*
[16:14:55] <ahuillet> heh, almost fit on a single line. :)
[16:16:41] <elibrokeit> Do you really have that issue with python "test" directories -- because that is broken packages
[16:17:17] <ahuillet> yes, test directories are present in the official arch packages of python
[16:17:20] <ahuillet> and they're not tiny
[16:17:22] <elibrokeit> And the rest is more or less just locales, heh
[16:17:43] <elibrokeit> Wait, oh, not "site-packages" then :p
[16:17:51] <ahuillet> what?
[16:17:53] <elibrokeit> Hmm....
[16:18:30] <elibrokeit> If site-packages/test/ exists it breaks the package guidelines because multiple packages cannot share the namespace!
[16:19:04] <ahuillet> FWIW I think the package that is tripping the one-step reinstall in my case is... man-pages.
[16:19:06] <ahuillet> no surprise there
[16:19:38] <elibrokeit> I.e. https://wiki.archlinux.org
[16:19:39] <phrik> Title: Python package guidelines - ArchWiki (at wiki.archlinux.org)
[16:19:53] <elibrokeit> ahuillet: man-pages can be fat, yes
[16:20:20] <ahuillet> fat man
[16:20:20] <elibrokeit> Maybe package a forked package "man-pages-fr" which provides/conflicts man-pages
[16:23:07] <ahuillet> oh I don't want man pages in French thank god
[16:25:22] <elibrokeit> Oh, I see -- just the locales :p
[16:25:33] <elibrokeit> So make man-pages-en :D
[16:26:00] <ahuillet> would Archlinux proper accept that at all?
[16:26:02] <ahuillet> oh you mean AUR
[16:29:11] <buildmaster> i686/adrdox is broken (says rechenknecht).
[16:36:42] <buildmaster> pentium4/linux-lts are broken (says rechenknecht).
[17:14:21] <buildmaster> i686/haskell-djinn-ghc is broken (says rechenknecht).
[17:27:28] <buildmaster> i686/haskell-base-compat-batteries are broken (says rechenknecht).
[17:29:40] -!- thePiGrepper has quit [Ping timeout: 246 seconds]
[17:32:45] -!- ahuillet has quit [Quit: ZNC - 1.6.0 - http://znc.in]
[17:34:36] -!- ahuillet has joined #archlinux32
[17:54:43] <buildmaster> i686/haskell-resolv is broken (says rechenknecht).
[18:00:19] -!- Tazmain has joined #archlinux32
[18:02:21] <Tazmain> Hi all, seems I haven't updated in a while, so now I keep getting that the packages are corrupt for all the packages. I downloaded: pacman, and the keyring from the web and installed with pacman -U, and tried to refresh the keys, but no dice
[18:21:43] <buildmaster> i686/perl-class-load-xs are broken (says rechenknecht).
[18:28:05] -!- eschwartz has joined #archlinux32
[19:00:24] -!- thePiGrepper has joined #archlinux32
[19:17:58] -!- City-busz has quit [Quit: http://quassel-irc.org - Csevegjen kényelmesen. Bárhol.]
[19:17:58] -!- elibrokeit has quit [Quit: A random quit message]
[19:18:16] -!- City-busz has joined #archlinux32
[19:18:19] -!- elibrokeit has joined #archlinux32
[19:30:13] <buildmaster> i686/haskell-network is broken (says eurobuild3).
[19:41:48] <buildmaster> i686/haskell-aeson is broken (says eurobuild3).
[19:49:51] <buildmaster> pentium4/haskell-aeson is broken (says eurobuild3).
[19:55:53] <buildmaster> i686/haskell-streaming-commons are broken (says eurobuild3).
[20:02:18] <buildmaster> i686/haskell-tasty-golden is broken (says eurobuild3).
[20:15:45] <buildmaster> i686/haskell-tls are broken (says eurobuild3).
[20:38:18] <buildmaster> i686/haskell-data-dword is broken (says eurobuild3).
[20:38:35] -!- ahuillet has parted #archlinux32
[20:45:27] <buildmaster> i686/haskell-spdx is broken (says eurobuild3).
[21:00:48] -!- thePiGrepper has quit [Ping timeout: 272 seconds]
[21:20:31] -!- yans has joined #archlinux32
[21:32:00] <girls> Tazmain: have you tried running `pacman-key --refresh-keys`?
[21:34:11] <Tazmain> girls: yep
[21:34:33] <girls> what is the exact error, pacman displays?
[21:35:03] <Tazmain> erm I can copy it, in a sec, its complaining about the pgp of the packjage, saying it corrupted
[21:36:40] <girls> the part about being corrupted is not really meaningful - it just means, that the signature was not accepted (for whatever reason was mentioned before)
[21:36:47] <Tazmain> package corrupted (PGP signature)
[21:37:57] <Tazmain> ok, i'll just run pacman-key again, and then run pacman with --debug
[21:37:58] <girls> that's all? nothing before that?
[21:39:01] <Tazmain> oh before it, it goes error: pacman signature X is marginal trust
[21:39:12] <girls> aha!
[21:39:41] <girls> what version of archlinux32-keyring do you have installed currently?
[21:39:54] <Tazmain> prob an old one, I can manually upgrade it
[21:41:55] -!- quequotion has quit [Quit: quequotion]
[21:44:04] <Tazmain> girls: well that was odd I was on the latest, but when i reinstalled it now, it worked
[21:44:08] <Tazmain> like I can upgrade
[21:45:00] <girls> `pacman-key --populate` probably did the trick - it's run in a post install hook
[21:45:09] <girls> I have no idea, though, why it was necessary
[21:45:25] <Tazmain> I did this pacman-key --refresh-keys again
[21:45:27] <Tazmain> a
[21:47:44] -!- thePiGrepper has joined #archlinux32
[21:57:25] -!- thePiGrepper has quit [Read error: Connection reset by peer]
[22:50:12] -!- thePiGrepper has joined #archlinux32
[22:59:58] <buildmaster> pentium4/perl-template-toolkit is broken (says eurobuild3).
[23:16:41] <buildmaster> pentium4/perl-async-interrupt is broken (says eurobuild3).
[23:47:57] -!- AndrevS has quit [Remote host closed the connection]