#archlinux32 | Logs for 2020-01-03
Back
[00:30:16] <buildmaster> i686/haskell-sbv is broken (says eurobuild6-3): https://archlinux32.org
[01:34:37] -!- isacdaavid has quit [Quit: Leaving.]
[01:53:42] -!- MrBIOS has quit [Quit: MrBIOS]
[02:03:34] <buildmaster> i686/hedgewars are broken (says eurobuild6-4): https://archlinux32.org
[02:25:48] <buildmaster> i686/shellcheck is broken (says eurobuild6-6): https://archlinux32.org
[02:36:20] -!- MrBIOS has joined #archlinux32
[03:09:44] -!- yans has joined #archlinux32
[03:46:59] <buildmaster> i486/armagetronad is broken (says nlopc46-i486bs0): https://archlinux32.org
[03:51:19] <buildmaster> i686/pandoc is broken (says eurobuild6-5) - already flagged out-of-date upstream on 2019-12-24: https://archlinux32.org
[04:34:23] -!- yans has quit [Ping timeout: 260 seconds]
[05:12:47] <buildmaster> i686/calibre is broken (says eurobuild6-1): https://archlinux32.org
[05:48:21] -!- MrBIOS has quit [Quit: MrBIOS]
[09:00:28] -!- deep42thought has joined #archlinux32
[09:00:29] <buildmaster> Hi deep42thought!
[09:00:29] <buildmaster> !rq deep42thought
[09:00:30] <phrik> buildmaster: <deep42thought> but let me help put the little terrorists to bed, first ...
[09:12:38] <buildmaster> i486/libarchive is broken (says nlopc46-i486bs0): https://archlinux32.org
[09:12:39] -!- abaumann has joined #archlinux32
[09:12:39] <buildmaster> Hi abaumann!
[09:12:39] <buildmaster> !rq abaumann
[09:12:40] <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. ;-)
[09:12:45] <deep42thought> Hi abaumann!
[09:12:53] <deep42thought> I see the first zst-related issues
[09:13:00] <deep42thought> do not use the current devtools (for now)
[09:13:17] <deep42thought> or replace all "zst" by "xz" in /usr/share/devtools32/makepkg*.conf
[09:18:45] <deep42thought> there are two possibilities to solve this (on my end): either we make the scripts suffix-agnostic (the right way) or we simply make the build scripts only accept zst-compressed packages from now on (the easy way) - which one do you prefer?
[09:19:04] <abaumann> hi deep42thought
[09:19:17] <abaumann> mmh.
[09:19:38] <abaumann> are we using scripts only with new zst packages?
[09:19:45] <abaumann> so everything we build from now on should be ok
[09:20:04] <abaumann> there are no effects on the end-user, right?
[09:20:06] <deep42thought> the new devtools will build zst by default
[09:20:10] <deep42thought> yes
[09:20:19] <deep42thought> the end-user should be unaffected
[09:20:30] <deep42thought> (at least, I saw mixed xz and zst packages upstream already)
[09:20:33] <abaumann> mmh. easy way first, if that fails, apply the more expensive method. :-)
[09:20:37] <deep42thought> ok
[09:20:39] <deep42thought> \o/
[09:31:30] <abaumann> a propos mirrors...
[09:31:35] <abaumann> ..we have a JSON report at https://www.archlinux32.org
[09:31:49] <deep42thought> yes, I know
[09:31:50] <abaumann> so, we could add it to the mirror status
[09:31:51] <deep42thought> I wrote it
[09:31:57] <abaumann> :-)
[09:32:01] <abaumann> the last_sync, I mean
[09:32:18] <abaumann> last_sync":"1970-01-01T00:00:00Z https://mex.mirror.archlinux32.org
[09:32:19] <abaumann> :-)
[09:32:22] <deep42thought> lol
[09:32:32] <abaumann> 2019-10-29T13:03:01Z
[09:32:46] <deep42thought> which mirror status page do you refer to?
[09:33:08] <deep42thought> this one? https://www.archlinux32.org
[09:33:09] <phrik> Title: Arch Linux 32 - Mirror Overview (at www.archlinux32.org)
[09:33:29] <abaumann> yes
[09:33:37] <abaumann> ah.
[09:33:42] <abaumann> that one is like upstreams
[09:34:06] <abaumann> ah
[09:34:06] <abaumann> https://www.archlinux.org
[09:34:08] <phrik> Title: Arch Linux - Mirror Status (at www.archlinux.org)
[09:34:23] <abaumann> this is the one upstreams
[09:34:39] <abaumann> though they measure things like completion..
[09:34:57] <deep42thought> well, we can display other columns - no problem
[09:35:48] <abaumann> I was thinking, we could redirect people to a nice page, so they are not complaining about stale packages or bugs, when they are just using a stale mirror
[09:38:59] <deep42thought> yeah, just print the info from json nicely formatted
[09:55:35] -!- MrBIOS has joined #archlinux32
[09:56:58] <buildmaster> i686/libarchive is broken (says eurobuild6-3): https://archlinux32.org
[09:58:34] <buildmaster> pentium4/libarchive is broken (says eurobuild6-6): https://archlinux32.org
[10:10:08] -!- AndrevS has joined #archlinux32
[10:33:38] <trotz> 2020/01/03 10:33 WARN buildmaster SMART sda (No output on stdout) stderr:
[10:33:38] <trotz> 2020/01/03 10:33 WARN buildmaster SMART sdb (No output on stdout) stderr:
[10:33:50] <abaumann> mmh
[10:36:37] <trotz> 2020/01/03 10:36 WARN buildmaster SMART sda (No output on stdout) stderr: >>> /etc/sudoers.d/nagios: syntax error near line 2 <<<
[10:36:44] <deep42thought> :-D
[10:37:06] <abaumann> another syntax I really hate
[10:38:06] <abaumann> WARNING: 24/25 OK - 1/25 WARNING
[10:38:07] <abaumann> aha
[10:38:26] <abaumann> so, the seagate disk is an older one
[10:39:05] <abaumann> or it's because it's a Seagate. :->
[10:39:10] <deep42thought> how original: /etc/makepkg.conf builds xz-compressed packages, devtools builds zst-compressed ones
[10:39:23] <abaumann> oh
[10:39:33] <deep42thought> we had massively failing seagate disks in the lab server (which hosts the master mirror)
[10:39:57] <abaumann> yeah, sorry to say that, but my experience goes into a similar direction..
[10:51:39] <trotz> 1970/01/01 01:00 OK buildmaster RAID
[10:51:56] <abaumann> 1.1.1970? not again.
[10:52:30] <deep42thought> you still remember the horrible old times?
[10:55:09] <abaumann> not really. I'm actually younger than that.. :-)
[10:55:56] <deep42thought> no offense
[10:56:02] <abaumann> none taken :-)
[10:57:42] * abaumann feels more secure with disk monitoring now
[10:58:16] <abaumann> did the same @work lately, upgraded the main company monitoring station, after 10 years of duty
[11:04:03] * abaumann prepares sort of a cabage cake ("Kohl-Waehe")
[11:04:28] <abaumann> (involves cutting cabbage into small pieces)
[12:08:31] <buildmaster> any/archlinux-menus are broken (says eurobuild6-3): https://archlinux32.org
[12:12:08] <deep42thought> the buildmaster *should* now only accept zst-compressed packages
[12:12:22] <deep42thought> the build slaves *should* also only build those (with current devtools32)
[12:12:47] <deep42thought> but we need to manually remove the cached build environments in /var/lib/archbuild/staging-*/
[12:13:08] <deep42thought> or alternatively only edit the /etc/makepkg.conf therein
[12:14:47] <abaumann> ah. ok.
[12:14:57] <abaumann> let me stop my slaves, they need a new-year cleaning anyway. :-)
[12:15:05] <deep42thought> !grab abaumann
[12:15:05] <phrik> deep42thought: Tada!
[12:20:51] <deep42thought> just great, namcap seems to choke on zst O.o
[12:21:23] <deep42thought> or is it my fault (again)?
[12:22:07] <abaumann> newer version of namcap needed?
[12:23:16] <deep42thought> no, namcap is not the problem
[12:23:26] <deep42thought> afk
[12:30:24] <abaumann> 5 days, 12:44:03 load avg: 1.53 running: test_socket (132 hour 29 min)
[12:30:27] <abaumann> python, of course
[12:30:52] <abaumann> testing != torture testing
[12:42:22] -!- thePiGrepper has quit [Ping timeout: 268 seconds]
[12:42:39] -!- thePiGrepper has joined #archlinux32
[12:52:47] <deep42thought> 5 days and the test slave did not yet starve? Give it another 5 days!
[12:57:32] <abaumann> mmh. eurbuild6-3, eurobuild6-5 and eurobuild6-6 are not running, but all of them are in post-build on https://archlinux32.org
[12:57:34] <phrik> Title: Arch Linux 32 - List of Build Slaves (at archlinux32.org)
[12:59:40] <deep42thought> they get removed after 1h
[12:59:58] <deep42thought> or you could actively report the build as aborted
[13:00:10] <deep42thought> ssh buildmaster return-assignment ABORT
[13:00:18] <deep42thought> with the correct ssh key for the respective build slave
[13:00:57] <abaumann> I have this touch /tmp/do-not-run-build-slave mechanism, maybe something is wrong there.. I'll check.
[13:01:26] <deep42thought> do you see a running ping-build-master process on the slave?
[13:01:29] <deep42thought> (you should not)
[13:01:50] <abaumann> only from slave2, and that's the python one.
[13:01:59] <deep42thought> ok, then it's fine
[13:02:04] <deep42thought> the buildmaster will kick you out in 1h
[13:02:23] <deep42thought> in case you want to abort early, you can employ the above ssh command
[13:02:29] <deep42thought> but IIRC, we do not do this automatically
[13:02:44] <deep42thought> we could/should do this probably when exiting ping-build-master ...
[13:02:53] <abaumann> I just wondered. It should actually finish the current build, then not start another one.
[13:03:10] <abaumann> Jan 03 13:02:44 eurobuild6 build-packages[561605]: I was too slow, the package is outdated. I will continue ...
[13:03:11] <deep42thought> probably the upload failed (due to zst vs xz)
[13:03:16] <deep42thought> exactly
[13:03:18] <abaumann> yep.
[13:03:21] <abaumann> ok then
[13:03:26] <deep42thought> the error message is somewhat misleading, though
[13:03:54] <abaumann> package compression doesn't change every week, so that's fine. :-)
[13:04:11] <deep42thought> hopefully :-D
[13:04:23] <deep42thought> btw: I think, it's still broken
[13:04:32] <deep42thought> so: expect more failures :-D
[13:09:00] <abaumann> do I have to update devtools32 on all slave?
[13:09:03] <abaumann> *slaves?
[13:09:11] <deep42thought> yes
[13:14:04] * abaumann declares eurbouild3 to be his update-the-slave guinea pig
[13:18:09] <abaumann> a local build produces which-2.21-5-pentium4.pkg.tar.zst
[13:18:11] <abaumann> this seems fine
[13:18:31] <deep42thought> the question is what the buildmaster does with that :-)
[13:18:53] <abaumann> next test. :-)
[13:22:30] <deep42thought> tar: Cowardly refusing to create an empty archive
[13:22:35] <deep42thought> *something* is still wrong :-D
[13:23:34] <deep42thought> `tar -J` won't extract packages anymore :-(
[13:32:18] <deep42thought> the next error will be a surprise (for me) - I think, it should work, now :-)
[13:33:48] <buildmaster> dirty! girls, my database - so dirty :-(
[13:33:48] * buildmaster goes insane.
[13:34:03] <deep42thought> :-D
[13:46:26] * buildmaster resumes sanity.
[13:48:04] <deep42thought> question: currently, the buildscripts dictate to build zst compressed packages (which is also the default of devtools32). Should rather the buildmaster dictate the build slave which compression to use? Or should the buildmaster take what it receives from the slave (e.g. the slave's configuration determines the compression)?
[13:58:09] <deep42thought> buildmaster: wtp python-scipy
[13:58:11] <buildmaster> deep42thought: i486/python-scipy: i486/community (1.2.1-1.1), i486/community-testing (1.3.1-2.1)
[13:58:11] <buildmaster> i686/python-scipy: i686/community-staging (1.4.1-1.0), i686/community (1.3.1-2.1)
[13:58:11] <buildmaster> pentium4/python-scipy: pentium4/community (1.3.1-2.1)
[13:58:23] <deep42thought> even this works
[13:58:24] <deep42thought> wow
[14:45:28] <abaumann> sorry, was short off
[14:46:49] <deep42thought> nothing to be sorry for - it looks, like everything works :-)
[14:47:09] <abaumann> ah. good. :-)
[14:50:36] <abaumann> sometimes, I am puzzled, just found eurobuild1 without any devtools installed
[14:50:45] <deep42thought> O.o
[14:51:03] <abaumann> ah. no.
[14:51:06] <abaumann> found it.
[14:51:12] <abaumann> wrong build user (build, not arch32)
[14:52:15] <abaumann> I didn't migrate it together with git.archlinux32.org
[15:03:05] <abaumann> I get tons of 'downloading ...*.zst'.. failed. Next ... at the end of a build..
[15:03:11] <abaumann> ok. with an any package.
[15:03:24] <abaumann> I think, I saw those messages before around namcap checking
[15:05:30] -!- davor has quit [Quit: Bye]
[15:07:48] -!- davor has joined #archlinux32
[15:14:27] <deep42thought> yes, the newly built packages try to compare themself to upstream packages - and those may not yet be zst
[15:14:52] <abaumann> ah, yes
[15:16:05] <deep42thought> we can make it try both .xz and .zst
[15:21:26] <deep42thought> ok, done
[15:25:48] <abaumann> This account has been terminated for violating Google's Terms of Service.
[15:26:00] <abaumann> cool. first ever video on youtube and it got terminated.
[15:26:29] <deep42thought> lol
[15:26:42] <abaumann> I wonder.
[15:27:11] <abaumann> so, apparently you cannot open accounts anymore not under your name.
[15:27:25] <abaumann> the video was basically a bootup-sequence of a Thinkpad 240X
[15:27:40] <deep42thought> I have to leave, cu
[15:27:42] <abaumann> and me opening a shell, compiling a Java program, opening Abiword.
[15:27:45] <abaumann> ok cu.
[15:27:53] <abaumann> I sulk to myself here ;-)
[15:28:05] -!- deep42thought has quit [Quit: Leaving.]
[15:32:40] <T`aZ> you probably violated some Oracle Java licenses or something
[15:33:00] * abaumann shrugs
[15:33:09] <T`aZ> did you run some static main() code ? i think they got the copyright for that ;)
[15:33:26] <abaumann> you're kidding, right? :-)
[15:33:42] <abaumann> should make them proud, 20 year old machine can serve as a developer machine for Java.
[15:33:51] <abaumann> not that Oracle has any merits to that.
[15:33:57] <abaumann> rather Sun.
[15:35:12] <T`aZ> yes im kidding, i have no idea, i tend to avoid google/broken products as much as possible :/
[15:35:23] <abaumann> I'm convined, GradeAUnterA couldn't open a Youtube acocunt, if he started one now. ;-)
[15:51:40] * T`aZ still wants a p2p, ipfs like network, but anonymous and that doesn't suck, we then would not need youtube anymore :p
[16:00:21] * abaumann nods in agreement
[16:09:50] -!- abaumann has quit [Quit: leaving]
[16:51:55] <bill-auger> T`aZ: thats what peertube is
[16:56:03] -!- c0ck4m0u53 has joined #archlinux32
[17:08:38] -!- davor has quit [Quit: Bye]
[17:10:25] -!- MrBIOS has quit [Quit: MrBIOS]
[17:14:17] -!- davor has joined #archlinux32
[17:14:17] -!- davor has quit [Remote host closed the connection]
[17:16:35] <buildmaster> i686/gitea is broken (says eurobuild6-5): https://archlinux32.org
[17:17:13] <buildmaster> i686/noise-repellent is broken (says eurobuild6-4): https://archlinux32.org
[17:17:40] <buildmaster> pentium4/gitea is broken (says eurobuild6-6): https://archlinux32.org
[17:19:27] -!- davor has joined #archlinux32
[17:22:02] <buildmaster> pentium4/noise-repellent is broken (says eurobuild6-3): https://archlinux32.org
[18:32:14] -!- abaumann has joined #archlinux32
[18:32:14] <buildmaster> Hi abaumann!
[18:32:14] <buildmaster> !rq abaumann
[18:32:15] <phrik> buildmaster: <abaumann> Welcome to the beatiful world of well-maintained Linux-software. :->
[18:32:54] <abaumann> deep42thought: I'm stopping my slaves, something is wrong: the number of to-be-build packages remains on 3023 though I get packages to build (and they run into outdated state)
[18:34:00] <abaumann> or maybe just the graph is wrong?
[18:34:59] -!- abaumann has quit [Client Quit]
[20:08:33] -!- deep42thought has joined #archlinux32
[20:08:34] <buildmaster> Hi deep42thought!
[20:08:34] <buildmaster> !rq deep42thought
[20:08:35] <phrik> buildmaster: <deep42thought> ahm, what is actually in the package "vulkan-headers" on i486? I thought, we disabled all vulcans?
[20:09:22] <deep42thought> abaumann: the two main questions are: 1. Does your build slave report any errors (in stdout or stderr) during the upload? 2. Does the built package appear on the master mirror?
[20:34:50] <deep42thought> abaumann: I see, that all returns of successfully-built packages have exit code != 0 in return-assignment - the error on your build slave will help me/you to find the underlaying problem
[20:35:04] <deep42thought> ^ that is only true for your build-slaves
[20:35:06] <deep42thought> mine work
[20:53:16] -!- MrBIOS has joined #archlinux32
[21:25:44] <buildmaster> i486/firetools are broken (says nlopc46-i486bs0): https://archlinux32.org
[21:30:13] <buildmaster> i486/libmysofa is broken (says nlopc46-i486bs1): https://archlinux32.org
[21:44:12] <deep42thought> I'll move xorgproto now to resolve the file conflict
[21:44:45] <deep42thought> hopefully, I do not break anything :-/
[21:46:51] * buildmaster goes insane.
[21:48:16] -!- thePiGrepper has quit [Ping timeout: 258 seconds]
[21:50:31] * buildmaster resumes sanity.
[22:15:00] -!- thePiGrepper has joined #archlinux32
[22:22:44] -!- deep42thought has quit [Quit: Leaving.]
[22:51:02] -!- c0ck4m0u53 has quit [Read error: Connection reset by peer]