#archlinux32 | Logs for 2019-07-31

Back
[00:50:53] -!- thePiGrepper has quit [Ping timeout: 245 seconds]
[03:24:48] -!- isacdaavid has quit [Quit: Leaving.]
[03:26:05] -!- thePiGrepper has joined #archlinux32
[04:48:47] -!- bill-auger has quit [Ping timeout: 245 seconds]
[04:53:25] -!- bill-auger has joined #archlinux32
[05:18:30] <buildmaster> pentium4/curaengine is broken (says eurobuild6-2): https://archlinux32.org
[05:19:01] <buildmaster> i686/curaengine is broken (says eurobuild6-1): https://archlinux32.org
[05:27:00] -!- isacdaavid has joined #archlinux32
[07:25:11] -!- thePiGrepper has quit [Ping timeout: 244 seconds]
[07:56:26] -!- isacdaavid has quit [Quit: Leaving.]
[08:36:49] -!- deep42thought has joined #archlinux32
[08:36:49] <buildmaster> Hi deep42thought!
[08:36:49] <buildmaster> !rq deep42thought
[08:36:50] <phrik> buildmaster: <deep42thought> windows suggests swap on external usb sticks :-)
[09:32:07] <deep42thought> buildmaster: what's up?
[09:32:08] <buildmaster> up? I'm up for 1 day, 1 hour, 39 minutes, load average: 1.23, 1.57, 1.61
[09:49:05] -!- thePiGrepper has joined #archlinux32
[10:52:23] -!- deep42thought has quit [Remote host closed the connection]
[10:55:14] -!- nit-picker has joined #archlinux32
[10:56:36] -!- deep42thought has joined #archlinux32
[10:56:36] <buildmaster> Hi deep42thought!
[10:56:36] <buildmaster> !rq deep42thought
[10:56:37] <phrik> buildmaster: * deep42thought listens carefully, but doesn't hear any bells ringing over here
[11:24:14] -!- nit-picker has quit [Remote host closed the connection]
[11:26:32] -!- nit-picker has joined #archlinux32
[11:56:10] -!- nit-picker has quit [Remote host closed the connection]
[12:19:09] -!- infides has joined #archlinux32
[12:26:13] <infides> Hey guys, I have installed archlinux32 on an old Fujitsu-Siemens Futro 400. Can someony help me to get the graphic card installed? It is an SiS adapter, so is it true, that only the lts-kernel works with it?
[13:39:21] -!- deep42thought has quit [Remote host closed the connection]
[13:39:39] -!- deep42thought has joined #archlinux32
[13:39:39] <buildmaster> Hi deep42thought!
[13:39:39] <buildmaster> !rq deep42thought
[13:39:40] <phrik> buildmaster: <deep42thought> my supervisor when I made my first shot with the laser: He handed me the remote and said: "Hey Erich, want to destroy a 20 Mio€ device?"
[13:56:28] -!- T`aZ has quit [Read error: Connection reset by peer]
[14:29:49] -!- deep42thought has quit [Quit: Leaving.]
[15:55:16] <buildmaster> i686/duktape is broken (says buildknecht2): https://archlinux32.org
[15:56:25] <buildmaster> pentium4/duktape is broken (says eurobuild6-1): https://archlinux32.org
[16:00:35] -!- AndrevS has joined #archlinux32
[16:01:03] -!- AndrevS has quit [Remote host closed the connection]
[16:10:57] -!- abaumann has joined #archlinux32
[16:10:58] <buildmaster> Hi abaumann!
[16:10:58] <buildmaster> !rq abaumann
[16:10:58] <phrik> buildmaster: <abaumann> stackoverflow is more or less a collection of esotheric modern woodoo instead of a source for technical information :-)
[16:16:46] <abaumann> infides: mmh, a SiS graphic chip, not sure if we have a package for video sis around (or even in the AUR)
[16:19:09] <abaumann> I'm actually quite surprised you could install a i686 version, and didn't have to resort to i486..
[16:19:32] <abaumann> https://aur.archlinux.org
[16:19:34] <phrik> Title: AUR (en) - xf86-video-sis (at aur.archlinux.org)
[16:20:28] <abaumann> You'll have to follow the links in the forums and you might be able to build it..
[16:20:39] <abaumann> ..there are rumours it got some patches for Xorg 1.20
[16:31:53] <infides> abaumann: thanks for the link, I will try it
[16:47:05] <abaumann> np
[16:47:09] <abaumann> good luck :-)
[16:49:21] -!- deep42thought has joined #archlinux32
[16:49:22] <buildmaster> Hi deep42thought!
[16:49:22] <buildmaster> !rq deep42thought
[16:49:23] <phrik> buildmaster: <deep42thought> so you're a garbage collector?
[16:54:27] <abaumann> hi deep42thought
[16:54:32] <deep42thought> Hi abaumann
[16:54:47] <deep42thought> the buildmaster fails quite a lot of the check-bugtracker mysql queries lately :-(
[16:55:10] <abaumann> so, there is a bug in the bugtracker? ;-)
[16:55:20] <deep42thought> no
[16:55:23] <deep42thought> mysql is overloaded
[16:55:30] <abaumann> mmh.
[16:57:12] * abaumann is scanning nice graphs in munin to see some performance hogs
[17:06:18] -!- infides has quit [Ping timeout: 245 seconds]
[17:08:36] -!- AndrevS has joined #archlinux32
[17:09:59] -!- infides has joined #archlinux32
[17:15:13] -!- infides has quit [Ping timeout: 244 seconds]
[17:17:15] <abaumann> mmh. bluez is broken in many ways..
[17:17:22] <abaumann> ..and blocks quite some packages..
[17:18:30] <deep42thought> btw: I'm fed up with rust
[17:18:39] <deep42thought> (have a look at the bugtracker for the latest state)
[17:20:04] <abaumann> yeah. fun.
[17:20:49] <abaumann> Rust will manage to kill linux userland, something Java never managed in 30 years :-)
[17:20:56] <deep42thought> !grab abaumann
[17:20:57] <phrik> deep42thought: Tada!
[17:21:27] <abaumann> this micro-build-system with cargo is the problem. rust the language per se is ok
[17:21:30] <abaumann> to be fair :-)
[17:21:42] <abaumann> though it has too much feature creep for my taste.
[17:21:52] <deep42thought> yes, it's really hard to patch anything
[17:22:07] <deep42thought> because it tries to do everything in a single "python ./x.py"
[17:22:12] <abaumann> yep.
[17:22:20] <abaumann> with downloading stage0 stuff
[17:22:48] <abaumann> and documentation is.. well bad.
[17:23:59] <abaumann> the worst part: I'm pretty sure that if when you start with a modern LLVM and bootstrap via mrust and from rust to rust, it will fail to. because it relies to heavily on the environment.
[17:24:13] <abaumann> *too
[17:24:26] * abaumann is spilling beer over all of his desk..
[17:24:38] * abaumann ..fetches something to clean the mess
[17:26:20] <abaumann> who is maintaining bluez? half of the tools are now orphaned (they got into AUR). I tried to create a ethernet connection latelty with a Nokia 112.. not a chance.
[17:26:52] <deep42thought> maybe upstream arch has too few maintainers, too?
[17:27:15] <abaumann> it's not their problem.
[17:27:19] <abaumann> I meant the original authors.
[17:27:24] <abaumann> the code looks.. well.. old.
[17:27:47] <abaumann> oh goodie: tests segfaulting and test-avdtp hangs on my build slave :-)
[17:28:09] * abaumann is happy about a somewhat less beerish working environment
[17:37:24] <abaumann> yeah: concerning rust: I got as far as trying to patch the N1024 errors somehow inside a stage, but the it failed again
[17:37:42] <deep42thought> you need to patch in build() and in package_rust()
[17:37:47] <deep42thought> only god knows, why
[18:08:13] -!- slacka123 has joined #archlinux32
[19:39:32] <abaumann> non262/Intl/NumberFormat/formatToParts.js:50:5 Error: Assertion failed: got "40", expected "41": value mismatch at 11
[19:39:38] <abaumann> in js60, almost funny :-)
[19:40:06] <deep42thought> find src -type f -exec sed -i s/41/40/g {} +
[19:40:36] <abaumann> sadly js60 has tons of rounding errors in the tests. I wonder, for a language this is quite unacceptable..
[19:40:51] <deep42thought> e^pi-pi=20
[19:42:02] <abaumann> e^(pi*i)+1=0 :-)
[19:42:11] <abaumann> if my rusty math brain is not mistaken.. :-)
[19:42:22] <deep42thought> that's correct
[19:42:33] <deep42thought> whereas e^pi-pi is only *almost* 20
[19:42:39] <abaumann> lol :-)
[19:42:49] <deep42thought> ... why should it be? ;-)
[19:44:25] <abaumann> ...physisists and their tendency to set everything to 1 and add a fancy named constant.. ;-)
[19:45:04] <deep42thought> 2pi=c=e=1
[19:45:19] <deep42thought> (speed of light, electron charge)
[19:45:56] <deep42thought> here are my sources: https://xkcd.com
[19:45:57] <phrik> Title: xkcd: e to the pi Minus pi (at xkcd.com)
[19:46:24] <abaumann> lol
[19:53:53] <deep42thought> well, in my field, "only" speed of light, wavelength, electron mass and electron charge are set to 1
[19:54:03] <deep42thought> ... which kind of makes sense ;-)
[19:54:17] <abaumann> it's about the relations..
[19:54:48] <abaumann> ..so, sort of hotfixed bluez for now.. to kick some packages out of the door.
[19:54:55] <deep42thought> \o/
[19:55:05] <deep42thought> kick them out of the door and down the stairs?
[19:55:31] <abaumann> if some packages break their neck, I wouldn't mind. :->
[19:56:17] <abaumann> js60 blocks 1074 packages?! wow.
[19:56:35] <abaumann> everything seems to embedd javascript nowadays..
[19:58:24] -!- isacdaavid has joined #archlinux32
[20:10:56] <deep42thought> the dependency can also be indirect
[20:11:01] <deep42thought> imagine meson depending on js60
[20:11:12] * buildmaster failed to execute a mysql query - can you have a look at "tmp.mysql-functions.query.2019-07-31T20:00:43.g6iaso.stdin"?.
[20:13:34] <deep42thought> though makedepends should not matter ... hmmm
[20:14:11] <abaumann> huh?
[20:14:15] * abaumann shudders
[20:14:33] <abaumann> not about the makedepends. but why to embedd a scripting language in a build tool..
[20:14:42] <abaumann> guile in make was already not the smartest idea.
[20:14:47] <abaumann> ..not to mention shell. ;-)
[20:20:52] <abaumann> dvd+rw-tools puzzles me: it fails in minor/major (missing sys/sysmacros.h) errors again.
[20:20:56] <abaumann> though it's fixed upstream
[20:21:00] <abaumann> ah.
[20:21:06] <abaumann> only in trunk *grmpf*
[20:21:24] <abaumann> https://git.archlinux.org
[20:21:26] <phrik> Title: dvd+rw-tools-7.1-minor-major.patch\trunk - svntogit/packages.git - Git clone of the 'packages' repository (at git.archlinux.org)
[20:21:39] <abaumann> who should I kick here?
[20:22:50] * abaumann considers summoning eli the wizard ;-)
[20:27:37] <deep42thought> abaumann: eli has no power over packages.git
[20:27:46] <deep42thought> consider adding a hotfix to our repo :-)
[20:27:59] <deep42thought> or simply building manually (there's a switch for bin/build-packages)
[20:28:21] <abaumann> okidoke
[20:35:40] <abaumann> buildmaster: why are you not sane? ;-)
[20:35:52] <deep42thought> pending mysql queries
[20:35:58] <abaumann> Lock wait timeout exceeded; try restarting transaction
[20:35:59] <deep42thought> I executed them manually
[20:36:04] <abaumann> that's worrying me a little bit
[20:36:12] <deep42thought> and am currently running a sanity-check
[20:36:16] <abaumann> ah. ok.
[20:36:38] <deep42thought> the frequency of the failed locks is concerning, though
[20:36:45] <abaumann> yup.
[20:39:00] <deep42thought> I wonder if that is the querie's fault or some hardware/performance issue
[20:40:15] <abaumann> mmh.
[20:41:03] <abaumann> mysql always on 100% and most likely a lot of stuff to do and some timeout set a little bit low.
[20:41:06] <abaumann> yeah.
[20:41:22] <deep42thought> note, that the queries are tried 10 times before finally failing
[20:41:46] <abaumann> oh
[20:42:18] <deep42thought> https://git.archlinux32.org
[20:42:19] <phrik> Title: archlinux32/builder: Tools for building 32-bit archlinux packages from archlinux.org's official, 64-bit tested PKGBUILDs et al. - Archlinux32 Gitea (at git.archlinux32.org)
[20:43:35] <abaumann> I was running some reschedules and prioritizations in parallel too..
[20:43:48] <abaumann> ..so maybe we just kicked mysql a little bit to much :-)
[20:44:03] * buildmaster resumes sanity.
[20:44:11] <abaumann> nice :-)
[20:44:19] * deep42thought dances the sanity dance
[20:44:37] <abaumann> a sane buildmaster make any day a happy day :-)
[20:44:41] <abaumann> *makes
[20:48:08] <deep42thought> what can be the cause of compute intensive mysql stuff?
[20:48:17] <abaumann> good question.
[20:48:23] <deep42thought> missing keys? too many keys? or simply too many / too complex queries?
[20:48:30] <abaumann> I lack decent monitoring tools..
[20:48:37] <deep42thought> I would imagine mysql maxing out memory
[20:48:44] <abaumann> tuning mysql sounds more like woodoo to me
[20:48:47] <deep42thought> ... but cpu???
[20:48:52] <abaumann> well.
[20:49:00] <abaumann> 100% cpu is not a bad sign per se.
[20:49:10] <abaumann> if the 100% is really the thing you have to do.
[20:49:16] <deep42thought> it means a faster cpu would accellerate stuff
[20:49:21] <abaumann> so 100% is actually bloddy efficient :-)
[20:49:26] <abaumann> true.
[20:49:46] <abaumann> I think we have some of those issues:
[20:49:59] <abaumann> restart mysql in shell looses all caching
[20:50:14] <deep42thought> when you're not doing time-critical stuff 100% cpu means you use the hardware properly, but when you're doing something "live", it means, you have no headroom left
[20:50:22] <abaumann> recursive queries are by nature very slow
[20:50:32] <deep42thought> what's recursive?
[20:50:38] <abaumann> dependencies mainly
[20:50:55] <abaumann> there is some hotspot maybe in statistics and ssh_log
[20:51:05] <abaumann> but that would be I/O intensive, not CPU intensive
[20:51:54] <abaumann> there is another bottleneck: updating package files (tar, xzing)
[20:52:08] <abaumann> you see it when a lot of haskell packages are built and all want to be uploaded
[20:52:23] <abaumann> this is not a problem, when bigger packages are being built
[20:52:31] <deep42thought> you think the actual repo-add is a bottleneck?
[20:52:49] <deep42thought> we could move the master mirror onto the buildmaster
[20:52:55] <abaumann> for sure recreating structures of 10'000 entries is not ideal, when just adding one package
[20:53:03] <deep42thought> this would get rid of the back-and-forth transfer of the *.db files
[20:53:08] <abaumann> yeah.
[20:53:20] <abaumann> but, is network really a problem at the moment?
[20:53:32] <deep42thought> you said so
[20:53:36] <abaumann> yeah.
[20:53:42] <deep42thought> or rather: I understood you that way
[20:53:50] <abaumann> but I think it's a degenerated case around small micro packages.
[20:54:01] <abaumann> so. not really something we should maybe bother.
[20:54:32] <abaumann> I don't see a big problem neither in build slaves, buildmaster performance currently, really..
[20:54:54] <abaumann> I fear we plateau sometimes because some crucial packages don't build and block all depenedencies.
[20:55:02] <deep42thought> yes
[20:55:04] <abaumann> as I wrote in the forum
[20:55:13] <abaumann> that's more dangerous to the project IMHO
[20:55:28] <deep42thought> all the other (infra) issues will pop up when we have >=4 architectures and/or many rebuilds
[20:55:38] <deep42thought> but it's mainly a "slow-down" not a "show-stopper"
[20:57:00] <abaumann> >=4 architecture? ;-)
[20:57:04] <abaumann> I wonder..
[20:57:29] <deep42thought> i386 i486 i4.586 i586 i686 pentium1 pentium2 pentium3 pentiumpi pentium4 pentium5 ...
[20:57:36] <abaumann> lol :-)
[20:58:01] <abaumann> what's nagging me more currently is: the rust, java issues are very tedious to debug and fix and have big impact.
[20:58:46] <abaumann> I like pentiumpi :-)
[20:59:22] <abaumann> though. If Intel designed it, it Pentium 3.1416 with a small bug somewhere..
[20:59:28] <deep42thought> yeah, the broken "low-level" packages are really annoying
[20:59:45] <deep42thought> pentiumpi = pentium3.1411251
[21:03:06] <abaumann> configure: error: Could not find required system headers; install the appropriate files from glibc-headers, libc6-dev and/or libattr-devel, libattr1-dev or use --disable-compile-against-syscalls if necessary.
[21:03:22] <abaumann> huh? java7-openjdk fails on i686 now suddenly too?
[21:04:20] <abaumann> oh. libdn is failing some tests..
[21:04:37] <abaumann> ..as pacman depends on it.. *cough*
[21:06:04] <abaumann> ../build-aux/test-driver: line 107: 18962 Aborted (core dumped) "$@" > $log_file 2>&1
[21:06:08] <abaumann> mmh. sounds sane.
[21:06:27] <abaumann> especially, because on 64-bit it's doing the same. :->
[21:06:53] <deep42thought> "only consider it a failure if it behaves different on x86_64"
[21:07:15] * abaumann throws a bug uptsream..
[21:11:31] * abaumann thinks git.archlinux.org is in some maintainance more right now..
[21:11:37] <abaumann> *mode
[21:12:25] <deep42thought> looks up and running to me
[21:12:32] <abaumann> mmh...
[21:12:46] <deep42thought> last commit to svn2git was 9 and 11 minutes ago
[21:13:16] <abaumann> I'm still fighting to get a systemd-netorkd with bridge working reliably
[21:14:38] <abaumann> same issue with asp32 on the buildmaster..
[21:14:42] <abaumann> asp32 -u export libidn
[21:14:44] <abaumann> just hangs
[21:15:22] <abaumann> fatal: unable to access 'https://git.archlinux.org/svntogit/packages.git/': The requested URL returned error: 504
[21:15:35] <abaumann> yeah. something is definitely happening over there..
[21:16:16] <deep42thought> git://git.archlinux.org works
[21:16:45] <abaumann> ok. now. it did.
[21:17:15] <abaumann> well.. I'm all in favour of maintaince.. and maintainance windows.. :-)
[21:23:24] * abaumann prepares a can with (very patriotic) sausages in a can for tomorrows first of August - nation day of Switzerland..
[21:24:49] <deep42thought> thuringian Bratwurst for the Swiss national holiday? :-D
[21:25:26] <abaumann> of course. :-)
[21:25:49] -!- arnd has joined #archlinux32
[21:31:20] -!- infides has joined #archlinux32
[22:01:12] -!- abaumann has quit [Quit: Lost terminal]
[22:32:52] -!- ofara_ has quit [Ping timeout: 258 seconds]
[22:34:42] -!- ofara_ has joined #archlinux32
[22:40:17] -!- ofara_ has quit [Remote host closed the connection]
[22:41:18] -!- ofara_ has joined #archlinux32
[22:54:11] <buildmaster> i686/squeak-vm is broken (says eurobuild6-6): https://archlinux32.org
[22:54:24] -!- deep42thought has quit [Quit: Leaving.]
[23:20:07] <buildmaster> pentium4/squeak-vm is broken (says eurobuild3): https://archlinux32.org
[23:37:32] -!- infides has quit [Ping timeout: 245 seconds]