#archlinux32 | Logs for 2018-04-06

Back
[00:37:42] -!- Mart|n has joined #archlinux32
[00:46:25] -!- davor has quit [Ping timeout: 265 seconds]
[00:52:33] -!- buildmaster has quit [Ping timeout: 264 seconds]
[00:53:15] -!- davor has joined #archlinux32
[01:16:02] -!- buildmaster has joined #archlinux32
[02:32:33] -!- isacdaavid has joined #archlinux32
[04:15:47] -!- Mart|n has quit [Remote host closed the connection]
[04:42:09] -!- oaken-source has joined #archlinux32
[05:02:30] -!- oaken-source has quit [Quit: Lost terminal]
[06:32:27] -!- davor has quit [Ping timeout: 240 seconds]
[06:38:08] -!- davor has joined #archlinux32
[07:20:38] -!- isacdaavid has quit [Ping timeout: 260 seconds]
[08:26:27] <girls> it looks, like systemd on the buildmaster crashed O.o
[08:27:53] <girls> "sudo reboot" -> "Failed to talk to init daemon."
[08:30:11] -!- buildmaster has quit [Ping timeout: 245 seconds]
[08:32:04] -!- buildmaster has joined #archlinux32
[09:28:36] -!- belanthor has joined #archlinux32
[09:31:05] * buildmaster failed to execute a mysql query - can you have a look at "tmp.mysql-functions.was_insane_query.stdin.2018-04-06T09:31:05.jxJmem"?.
[10:04:57] -!- eduardoeae has quit [Ping timeout: 240 seconds]
[10:10:08] -!- abaumann has joined #archlinux32
[10:10:21] <abaumann> buildermaster says: "Connection to mysql database failed: No such file or directory"
[10:12:21] <abaumann> spam in the forum is still a problem.
[10:17:19] -!- abaumann has quit [Quit: leaving]
[12:10:07] -!- eduardoeae has joined #archlinux32
[12:17:57] <girls> abaumann: I've noticed both, but have no real handle to either of them ...
[13:07:15] -!- deep42thought has joined #archlinux32
[13:07:43] <deep42thought> abaumann: If you can provide me a criterium for spam, I can implement it in the forum
[13:24:12] -!- abaumann has joined #archlinux32
[13:25:04] <abaumann> deep42thought: as you suggested.. a tool showing latest users and their posts together with a simple "delete-all-from-that-user" would be nice.
[13:25:38] <deep42thought> there is already a "delete user including all his/her posts"
[13:25:49] <abaumann> huh? in the interface?
[13:25:58] <deep42thought> maybe, we should promote you to "admin"
[13:26:01] <deep42thought> yes
[13:26:17] <abaumann> ah. :-)
[13:27:27] <deep42thought> do you see it now?
[13:28:01] <abaumann> ah..
[13:28:07] <abaumann> this looks very different. :-)
[13:28:09] <abaumann> thanks.
[13:28:43] <abaumann> ..and now there is no spam.. *snif8
[13:30:16] <deep42thought> sry, I deleted that one already ...
[13:31:30] <deep42thought> and regarding the buildmaster: mysqld crashes and I'm currently looking into stack traces ...
[13:32:39] <City-busz> deep42thought: why are chromium and qt5-base still in staging?
[13:33:32] <abaumann> mysqld crashes? Is it running on Arch or Arch32?
[13:33:44] <deep42thought> archlinux
[13:33:45] <abaumann> I had some really weird issues with segfaults this morning with OSC/python2
[13:33:58] <abaumann> OpenSuse Build Service command line tool.
[13:34:08] <abaumann> also on arch..
[13:34:15] <deep42thought> :-/
[13:34:28] <deep42thought> City-busz: I cannot really comment on that when mysqld does not run
[13:34:53] <abaumann> I'm really in favour or rebuilding everything after an upgrade in the toolchain..
[13:34:58] <abaumann> *of
[13:35:13] <deep42thought> but if my logic works as intended, then something in [testing] will be replaced by the stuff currently stuck in [staging] which is still needed by something in [testing]
[13:38:52] -!- eduardoeae has quit [Remote host closed the connection]
[13:40:26] -!- eduardoeae has joined #archlinux32
[13:43:19] <City-busz> chromium and qt5-base in [staging] are linked to icu-61, but icu-61 is already in [testing]
[13:43:36] <deep42thought> that was the reason, why I rebuilt qt5-base
[13:43:48] <deep42thought> but I cannot figure out why it is stuck in [staging]
[13:43:51] <City-busz> so I had to install chromium and qt5-base from [staging] to avoid missing library
[13:44:40] <abaumann> icu-61 was also a problem in ArchlinuxARM, in postfix.
[13:45:49] <City-busz> so the question is: how can it happened that icu-61 is in [testing] while the rebuilt chromium and qt5-base are still in [staging] :)
[13:46:37] <deep42thought> chromium is a mistake, but qt5-base was simply scheduled after icu-61 was already in testing
[13:50:00] -!- esotericnonsense has quit [Read error: Connection reset by peer]
[13:50:26] -!- esotericnonsens_ has joined #archlinux32
[13:51:42] -!- abaumann has quit [Quit: leaving]
[13:53:38] <deep42thought> ah, it looks like the job for qt5-base was created before the one for chromium, so this part fits, too
[13:53:58] <deep42thought> maybe I rescheduled chromium and forgot about it :-/
[14:16:24] -!- abaumann has joined #archlinux32
[14:16:45] <abaumann> mmh. my osc problem seems to be a weird one. my test machine is probably borked..
[14:17:58] <deep42thought> https://ptpb.pw
[14:20:58] <abaumann> InnoDB: Failing assertion: addr.page == FIL_NULL || addr.boffset >= FIL_PAGE_DATA
[14:21:09] <abaumann> memory corruption, library corruption, etc.
[14:21:19] <abaumann> really hard to find..
[14:26:12] <deep42thought> ok, we got a backup from 2 hours before the crash - I'll simply use that :-)
[14:27:00] <deep42thought> ... or I could try to use the mirrored database, but I have no idea how to do that
[14:28:22] <abaumann> I would also go back to the backup.
[14:28:37] <abaumann> To me it seems the database is just corrupt.
[14:28:50] <deep42thought> yes
[14:28:55] <deep42thought> question is: why
[14:28:59] <abaumann> I hope the backup is some sort of SQLish..
[14:29:11] <abaumann> crash of the vm?
[14:33:33] <deep42thought> ah, no: the backup is from after the crash
[14:33:42] <deep42thought> (there was a reboot after the crash)
[14:33:53] <deep42thought> last line of mariadb's log before the crash: "If you see no messages after this, something went terribly wrong..."
[14:33:57] <deep42thought> :-/
[14:34:25] <abaumann> yeah. mysql is sort of a toy. not a real database.. never liked it.
[14:34:35] <abaumann> you can force the startup of innodb.
[14:34:47] <deep42thought> read that, too
[14:34:51] <deep42thought> I'll give it a shot
[14:35:03] <abaumann> so maybe you can get a sql dump this way on the recovered database
[14:37:28] <abaumann> ok. the python error with osc happens also on a competly different archlinux 64-bit machine.
[14:37:43] <abaumann> it's a segfault in the memory management while doing digest stuff in openssl
[14:37:46] <deep42thought> sounds like it's time for a bug report, then
[14:38:00] <abaumann> a real bug report :-)
[14:38:26] <deep42thought> :-)
[14:38:34] <abaumann> I want to find out, what causes it.. prefereably also with a simple python script. installing osc to reproduce the bug seems a little bit heavy..
[14:40:22] esotericnonsens_ is now known as esotericnonsense
[14:48:51] <abaumann> aha: openssl 1.1.0h crashes, 1.1.0g works.
[14:49:10] <abaumann> and difflog shows me an upgrade on 28th of March..
[14:51:05] <abaumann> more programs should crash, I wonder. Or osc uses a very obscure/old feature in openssl causing the crash.
[14:51:30] <deep42thought> maybe something, that got deprecated and was removed?
[14:52:09] <abaumann> 1.1.0g to 1.1.0h? well, we speak openssl here :-)
[14:52:43] <abaumann> if I remember correctly, it's an old python2 module m2crypto.
[14:52:54] <abaumann> so it could easily be broken..
[14:54:37] <abaumann> or maybe an ABI bump in m2crypto. I'll try to rebuild that one against openssl 1.1.0h..
[14:55:29] <deep42thought> whoah, the dumped tables seem to be clean :-D
[14:56:11] * buildmaster resumes sanity.
[14:56:55] <abaumann> good :-)
[14:57:18] <deep42thought> now I just need to check, how the database mirror got through the restore procedure ...
[14:58:33] <abaumann> I should really start reading bugreports: https://bugs.archlinux.org
[14:59:07] <deep42thought> this is already 3 days old - one can easily miss that
[14:59:13] <deep42thought> ;-)
[14:59:29] <abaumann> I'm glad, it's only an isolated python2 issue..
[15:27:21] <tyzoid> deep42thought: Can you do a mysqldump from the slave db?
[15:28:20] <deep42thought> tyzoid: haven't tested
[15:28:27] <deep42thought> but it's running again, already
[15:28:32] <tyzoid> nice.
[15:28:35] <deep42thought> recovery level 2 worked
[15:28:54] <tyzoid> Sweet. Might be a good idea to review the slave restore procedure, though
[15:29:18] <tyzoid> Not necessarily now, though, just at some point.
[16:15:27] -!- Alina-malina has quit [Ping timeout: 256 seconds]
[16:56:26] <buildmaster> pymol is broken (says buildknecht3).
[17:31:20] <deep42thought> I think, documenting the restore procedure might be enough
[17:33:14] -!- abaumann has quit [Remote host closed the connection]
[17:38:15] -!- isacdaavid has joined #archlinux32
[17:38:16] -!- isacdaavid has quit [Remote host closed the connection]
[17:43:31] -!- isacdaavid has joined #archlinux32
[17:47:13] -!- davor has quit [Ping timeout: 265 seconds]
[17:52:47] -!- davor has joined #archlinux32
[18:07:35] -!- belanthor has quit [Quit: Leaving]
[18:11:30] -!- Mart|n has joined #archlinux32
[18:58:25] <deep42thought> tyzoid: It looks, like resuming the master from a slave mysqldump is non-trivial
[18:58:41] <tyzoid> hmm
[18:58:43] <deep42thought> maybe, we should use a multi-master setup?
[18:58:48] <tyzoid> perhaps
[18:59:03] <tyzoid> but IIRC multi-master cross-continent might have issues
[18:59:19] <tyzoid> maybe that question is suited for the #mariadb or #mysql channels?
[18:59:31] <deep42thought> good idea
[19:00:06] <tyzoid> whoops, the mariadb channel is #maria, not #mariadb
[19:00:33] <deep42thought> :-D
[19:00:58] <deep42thought> "17.7.2 Group Replication Limitations" does not list latency
[19:01:34] <tyzoid> Even so, we need the transactions to be serializable on all masters.
[19:01:34] <deep42thought> ah, "17.7.1 Group Replication Requirements" does
[19:01:43] <tyzoid> yup
[19:01:44] <deep42thought> "Network Performance.  Group Replication is designed to be deployed in a cluster environment where server instances are very close to each other, and is impacted by both network latency as well as network bandwidth. "
[19:02:04] <tyzoid> mostly because strict ordering is necessary
[19:02:40] <tyzoid> performance is limited because iirc it votes
[19:03:03] <deep42thought> I think, I'll try the steps in that tutorial this evening in your vm on a separate mariadb instance
[19:03:20] <deep42thought> so I can test cross-atlantic group replication between your vm and the master mirror
[19:03:28] <tyzoid> I can clone the container and give you another one, if you want
[19:03:55] <deep42thought> yeah, might be cleaner
[19:04:04] <tyzoid> though port forwarding might be an issue. What ports would you like?
[19:04:46] <deep42thought> any two are fine
[19:04:49] <deep42thought> ipv4
[19:05:03] <deep42thought> "The group communication engine used by MySQL Group Replication only supports IPv4. Therefore, Group Replication requires an IPv4 network infrastructure."
[19:07:36] <tyzoid> ok, container is cloning
[19:18:07] <tyzoid> deep42thought: srv1.tyzoid.com -p 2205
[19:18:22] <tyzoid> port 3307 is forwarded to 3306 on the container, port 33071 is forwarded to 33061 on the container.
[19:18:56] <deep42thought> "No route to host"
[19:19:59] <tyzoid> Alright, try now
[19:20:06] <tyzoid> looks like the IP change on the guest didn't hold
[19:20:46] <deep42thought> yep, looks good
[19:20:47] <deep42thought> thx
[19:21:14] <tyzoid> No prob. I forwarded ipv6 and ipv4 on all three ports for convenience
[19:21:23] <tyzoid> probably not necessary, but I like things consistant
[19:22:00] <tyzoid> lol, at this rate, I'll be needing another hypervisor machine
[19:22:14] * tyzoid thinks about moving srv0 to srv1, reprovisioning srv0, then moving back :P
[19:24:29] <deep42thought> well, I'd say, _that_ vm is really temporary just for this test
[19:45:55] <tyzoid> Not a problem, just that my other machine has 2x the ram
[19:58:16] <tyzoid> hey deep42thought: Can you help me take a look at srv1-sql-tests?
[19:58:28] <tyzoid> ssh -p2203?
[19:59:42] <deep42thought> yes?
[20:00:37] <tyzoid> run htop
[20:00:46] <tyzoid> the memory usage makes little sense to me
[20:00:59] <deep42thought> top reports 18%
[20:01:06] <deep42thought> which makes more sense
[20:01:47] <tyzoid> right, but free show something different too
[20:01:58] <tyzoid> htop shows 18% is for the mysql process
[20:02:04] <deep42thought> ah, right
[20:02:37] <tyzoid> I don't see any large allocations in tmpfs either
[20:02:59] <deep42thought> yeah, that's strange
[20:03:05] <deep42thought> I have no idea ...
[20:03:37] <tyzoid> Part of me trying to debug why the host shows 7/8 gb used, but only only 4gb memory used expected
[20:04:21] <tyzoid> so unless the 18% figure is of the hosts's memory, it doesn't make much sense
[20:04:44] <tyzoid> mysql 3420 0.1 18.0 947120 189544 ? Ssl Apr02 6:44 /usr/bin/mysqld
[20:04:56] <tyzoid> it's showing 947,120 MB VSZ, but only 189544 RSS
[20:05:07] <tyzoid> and 18% would be the rss
[20:05:20] <tyzoid> which would suggest it's of the 1g ram allocation
[20:06:51] <tyzoid> Time for me to ask in #archlinux
[20:12:21] <tyzoid> deep42thought: Did you ask in #maria?
[20:12:30] <tyzoid> re. the multimaster?
[20:15:27] <deep42thought> tyzoid: nope
[20:15:38] <deep42thought> I thought, I'd just try it :-)
[20:16:05] <deep42thought> plus: I had to chill down a screaming child
[20:20:03] <deep42thought> re RAM: I had created a mysqldump in /tmp/, copied it and deleted it - maybe it's still there because someone (who?) keeps it open?
[20:20:18] <deep42thought> I'm not sure, how big it was, though
[20:21:03] <deep42thought> 15MB - that's probably not the cause
[20:23:11] <tyzoid> Yeah, possible but unlikely
[20:23:22] <tyzoid> plus the kernel would (should?) show that as available anyway
[20:55:11] -!- chub has joined #archlinux32
[20:56:56] <chub> hi guys, me again, I've updated my arch linux 32bit machine and it was now time to recompile my custom usb ethernet module for asix. I have to stick to a particular version of the driver as my device is one of those "clones" and not an authentic device, but the older version of the driver was working fine. I started using the older version because the kernel's built-in version stopped working all together, same as with the newer version of the driver
[20:57:11] <chub> so now after I rebooted I tried to compile it but I'm running into errors
[20:57:26] <chub> and I'm not a driver developer so I have no idea how to resolve this issue
[21:03:14] -!- abaumann has joined #archlinux32
[21:06:02] <tyzoid> chub: Got a log?
[21:06:33] <tyzoid> deep42thought: Want to post the #maria log on mirror.archlinux32.org?
[21:06:47] <tyzoid> got permission from the two guys that were included
[21:07:09] <tyzoid> I think it'd be a handy reference to keep around
[21:07:16] <deep42thought> yeah, ok
[21:07:33] <chub> ok, I guess I need to paste it somewhere
[21:07:50] <tyzoid> chub: curl -F c=@- https://ptpb.pw
[21:07:58] <tyzoid> you can pipe your log to that
[21:08:05] <tyzoid> or curl -F c=@/path/to/logfile.log https://ptpb.pw
[21:10:16] <chub> https://pastebin.com
[21:10:23] <chub> sorry I only read the message now
[21:10:40] <tyzoid> deep42thought: Already scrubbed; https://dl.tyzoid.com
[21:10:53] <tyzoid> might take a bit to reformat, but that should be what we want to post
[21:10:59] <tyzoid> chub: please use something other than pastebin
[21:11:47] <chub> ok I'll try ptpb.pw
[21:11:56] <tyzoid> deep42thought: If you'd prefer, I can put it up
[21:12:33] <deep42thought> no, I'm already formatting it :-)
[21:13:23] <chub> https://ptpb.pw
[21:13:24] <abaumann> chub: mmh. is there a -Werror when building?
[21:13:58] <chub> abaumann, yes I've tried disabling warnings, like adding -w but that didn't help
[21:15:05] <abaumann> mmh.. trying to build asix-module myself..
[21:15:50] <abaumann> ui. the dependency list of the module reads like a big book. :-)
[21:18:56] <chub> there used to be a asix-module, but the version on arch was too high so I have to stick to my older version
[21:19:54] <chub> or at least, had to
[21:22:37] <tyzoid> chub: Btw, we do have old versions of arch32 packages on https://archive.archlinux32.org
[21:22:59] <tyzoid> I don't think we've got your module on there from far enough back to be helpful
[21:23:10] <tyzoid> but if there's a kernel issue, it might help testing on an older kernle
[21:23:14] <tyzoid> kernel*
[21:23:15] <abaumann> ..and it has to fit to the kernel..
[21:23:26] <abaumann> https://pastebin.com really scarry internet. :-)
[21:23:35] <abaumann> https://aur.archlinux.org
[21:23:56] <abaumann> also has a message with the same error. I'm afraid internal kernel interfaces changed and the driver has to be adapted..
[21:24:01] <tyzoid> abaumann: Pastebin has been deleted?
[21:24:45] <deep42thought> delete the colon
[21:24:51] <abaumann> yep. sorry.
[21:24:58] <abaumann> It's the same error message..
[21:25:00] <tyzoid> ah, lol
[21:25:04] <abaumann> :-)
[21:25:40] <chub> yes, I was about to say
[21:25:50] <chub> dev->delay.data = (unsigned long) dev;
[21:25:57] <chub> that seems to be the main problem
[21:26:05] <abaumann> dev->delay.data = (unsigned long) dev;
[21:26:19] <abaumann> urgh. I feel 32-bit problems creeping up..
[21:27:12] <chub> ok I guess I'll have to watch asix-module or something
[21:27:53] <chub> tyzoid, would it make more sense then to downgrade my kernel?
[21:28:05] <tyzoid> chub: Was it working on an older kernel?
[21:28:43] <chub> tyzoid, yes
[21:28:47] <tyzoid> Downgrading the kernel is just a temporary bandaid. If it works, it doesn't actually fix the problem, since you'll need to upgrade it later.
[21:29:07] <tyzoid> I'd say if you want it to work while you try to fix it on the newer kernel, than that'd be fine
[21:29:09] <chub> yes, I understand, my wifi is connected to this usb ethernet dongle... kinda need it
[21:32:00] <abaumann> linux 4.15.15-1.0
[21:32:00] <abaumann> linux-api-headers 4.14.8-1.0
[21:32:06] <abaumann> and a timer interface update.
[21:32:46] <abaumann> are those supposed to be out of sync?
[21:33:22] <chub> thanks everyone for trying to help me
[21:33:44] <deep42thought> abaumann: they're not
[21:34:17] <abaumann> well, they are upstream. :-)
[21:34:45] <abaumann> linux-api-headers should be used against glibc. kernel modules I think are best compiled against the sources of the running kernel..
[21:34:56] <deep42thought> also I cannot see linux-api-headers in a newer version somewhere in our repos
[21:35:55] <chub> alright here goes nothing, rebooting into older kernel. first time for me to do a downgrade
[21:36:02] -!- chub has quit [Quit: Leaving]
[21:36:19] <abaumann> well.. most likely has nothing to do with it..
[21:36:59] <abaumann> https://lkml.org
[21:36:59] <tyzoid> abaumann: looks like we've got linux 4.15.6-1.0 and linux-api-headers 4.14.8-1.0
[21:37:32] <abaumann> seems ok.
[21:37:49] <abaumann> I think, the kernel Makefile uses the build dir in /usr/lib/modules/4.15.15-1.0-ARCH/build with local header files.
[21:37:56] <abaumann> the header files in api-headers are for userland.
[21:38:04] <tyzoid> iirc that's correct
[21:38:15] <tyzoid> doesn't really make sense any other wya
[21:38:17] <abaumann> seems, there was a kernel update in 4.15 regarding timers..
[21:38:19] <tyzoid> way*
[21:38:32] <abaumann> ^ link above.
[21:40:11] <tyzoid> abaumann: looks like people in x86_64-land are complaining about the same compilation error with asix: https://aur.archlinux.org
[21:40:35] <abaumann> yep.
[21:41:00] <abaumann> usually the trick is to read code of other similar modules and see what has to be adapted..
[21:41:18] <abaumann> .. that said, the original author is better of to do the change. :-)
[21:45:55] <tyzoid> wow, pacman -Fs is *slow*
[21:47:11] <deep42thought> tyzoid: that's part of the reason, why "wtf" is so slow on the buildmaster
[21:47:26] <tyzoid> deep42thought: Yeah, that's why I was trying to speed it up
[21:47:37] <tyzoid> but even putting the files database in tmpfs didn't help much
[21:49:01] <abaumann> https://lwn.net
[21:49:08] -!- chub has joined #archlinux32
[21:49:16] <tyzoid> wb chub
[21:49:32] <chub> hello, just wanted to report, that my module is working on older 4.14 kernel
[21:49:33] <tyzoid> chub: looks like people in x86_64-land are complaining about the same compilation error with asix: https://aur.archlinux.org
[21:49:36] <chub> thanks tyzoid
[21:49:52] <abaumann> my only question is, the setup_timer says flags as parameters and data in the article, so how can I pass the custom data to the callback *puzzle*
[21:50:22] <abaumann> " For the time being, timer_setup() simply stores a pointer to timer in the data field. Note that the prototype of the callback has changed to expect the timer_list pointer. "
[21:50:30] <abaumann> well, so I could ommit the ->data assignment.
[21:50:34] <chub> 4.14.15-1.0-ARCH #1 SMP PREEMPT Fri Jan 26 00:21:11 CET 2018 i686 GNU/Linux
[21:50:38] <abaumann> I love changing code I cannot possibly test. :-)
[21:50:49] <chub> ok have a good evening, thanks everyone
[21:50:59] -!- chub has quit [Client Quit]
[21:52:34] <abaumann> dev->delay.data = (unsigned long) dev;, yeah, that's exacltly the thing kernel developers don't want. :-)
[21:54:06] <tyzoid> abaumann: wait what?
[21:54:16] <tyzoid> just realized that they're converting the pointer and assigning it
[21:54:30] <abaumann> yeah.. that's why the interface was changed. :-)
[21:56:14] * buildmaster goes insane.
[21:57:21] <deep42thought> hmm
[21:57:29] <deep42thought> it sees pending mysql queries
[21:57:44] <deep42thought> but actually, all queries should lock the sanity check before
[21:57:55] <abaumann> [mkarchiso] INFO: Done! | 471M /build/arch/archlinux-2018.04.06-i486.iso
[21:57:58] <abaumann> sweet. :-)
[21:57:59] * buildmaster resumes sanity.
[21:58:21] <deep42thought> ui
[21:58:25] <deep42thought> nice
[21:58:53] <deep42thought> btw: the database is not really dual-arch ready :-(
[21:58:54] <abaumann> Now, let's see if I can mount the thing via PXE/TFTP/NBD.
[21:58:59] <abaumann> no worries. :-)
[21:59:15] <abaumann> I'll have to make another trip around the block with all the stages..
[21:59:21] <abaumann> ..quite a mess currently. :-)
[21:59:53] <tyzoid> deep42thought: https://pkgapi.arch32.tyzoid.com
[22:00:06] <tyzoid> You can query json, and it'd make WTF faster
[22:00:30] <abaumann> nice. and fast.
[22:00:40] <tyzoid> abaumann: is that built with my iso script?
[22:01:09] <abaumann> partially: I use archiso32-git and snippets from your releng script
[22:01:15] <deep42thought> tyzoid: are the provided files also available via json?
[22:01:28] <abaumann> mainly the PGP key and the mirror stuff.
[22:01:30] <tyzoid> it uses pkgfile to generate the list
[22:01:41] <tyzoid> pacman -Fs does, but not pkgfile, afaik
[22:01:43] <tyzoid> I can check the manpage
[22:03:16] <tyzoid> looks like yes. Let me update it to handle this
[22:09:40] <tyzoid> deep42thought: https://pkgapi.arch32.tyzoid.com
[22:10:05] <tyzoid> I can construct a more structured object, by splitting out the version into an info block
[22:10:06] <tyzoid> if you want
[22:11:05] <deep42thought> yes, might be nice
[22:13:09] <abaumann> no /new_root/sbin/init.. mmmh.
[22:15:06] <abaumann> no sysv compat for systemd.. weird. should be part of pacstrap.. or of packages.both.
[22:18:33] -!- abaumann has quit [Quit: leaving]
[22:19:46] <deep42thought> tyzoid: If you want / have time, you can modify "wtf" from our build scripts to work with your json interface
[22:20:13] <deep42thought> it's called by the buildmaster's irc bot
[22:20:33] <tyzoid> perhaps later. going to finish this bit up. Do remind on saturday, though, since I'll likely forget
[22:20:46] <deep42thought> :-D
[22:20:50] <deep42thought> I'll try
[22:20:55] <deep42thought> I'll go to bed now
[22:20:59] -!- deep42thought has quit [Quit: Leaving.]
[22:21:00] <tyzoid> ok, g'night
[22:22:23] -!- abaumann has joined #archlinux32
[22:22:31] <tyzoid> wb abaumann
[22:22:39] <abaumann> ah. some booting. :-)
[22:30:53] <abaumann> oh. silly me. I really forgot to build a 486 systemd:
[22:30:54] <abaumann> # we don't need systemd-sysvcompat
[22:30:54] <abaumann> #sed -i "/pkgname=/,/)/ s/'systemd-sysvcompat'//" PKGBUILD
[22:31:02] <tyzoid> lol
[22:31:05] <abaumann> there was no comment there of course.
[22:32:31] <abaumann> On the other hand: both the ISO and the HD installation boot fine just with init=/lib/systemd/systemd without the sysvcompat package.
[22:32:58] <tyzoid> so... what packages need sysvcompat?
[22:33:03] <abaumann> none.
[22:33:11] <abaumann> just the archiso builder..
[22:33:27] <tyzoid> hmm
[22:33:36] <abaumann> or better put, the syslinux.cfg assumes that init is /sbin/init (which is a symlink in sysvcompat, I believe)
[22:36:00] <tyzoid> why not just symlink /sbin/init to /lib/systemd/systemd
[22:36:26] <tyzoid> call the package systemd-init-shim
[22:36:27] * buildmaster goes insane.
[22:36:34] <tyzoid> lol, buildmaster didn't like my idea
[22:39:44] <tyzoid> abaumann: https://pkgapi.arch32.tyzoid.com
[22:39:46] <tyzoid> any thoughts?
[22:39:54] <tyzoid> xml version also available: https://pkgapi.arch32.tyzoid.com
[22:40:02] <tyzoid> https://pkgapi.arch32.tyzoid.com <-- whoops, this is the correct link
[22:41:20] <abaumann> ah. the '\/' in JSON was confusing me. but if yoy embedd it into Javascript, escaping it makes sense, I assume.
[22:41:30] <tyzoid> yup, that's just the json encoding
[22:41:37] <tyzoid> the xml doesn't have that
[22:41:57] <tyzoid> but yeah, the api speaks both json and xml
[22:42:01] <tyzoid> ex: https://pkgapi.arch32.tyzoid.com
[22:42:10] <tyzoid> and: https://pkgapi.arch32.tyzoid.com
[22:42:32] <tyzoid> you could use .json as the extension, but it doesn't change anything, since json is default
[22:44:35] <tyzoid> Here's a more complex test case: https://pkgapi.arch32.tyzoid.com / https://pkgapi.arch32.tyzoid.com
[22:47:03] <abaumann> Didn't know there where so many watches in Linux ;-)
[22:47:22] <abaumann> more than 'clock'..
[22:49:16] <tyzoid> yup. Anyway, I'm going to look at patching deep42thought's buildmaster irc bot to use this sometime tomorrow
[22:50:15] <abaumann> yeah. good idea.
[22:50:46] <tyzoid> buildmaster: wtf watch
[22:51:09] <tyzoid> this might take some time
[22:51:29] <buildmaster> testing/bash-completion
[22:51:29] <buildmaster> testing/breeze
[22:51:29] <buildmaster> testing/oxygen
[22:51:29] <buildmaster> testing/xcursor-themes
[22:51:29] <buildmaster> community-testing/deepin-icon-theme
[22:51:29] <buildmaster> ... (24 lines total)
[22:51:29] <buildmaster> testing/bash-completion
[22:51:30] <buildmaster> testing/breeze
[22:51:30] <buildmaster> testing/oxygen
[22:51:31] <buildmaster> testing/xcursor-themes
[22:51:31] <buildmaster> community-testing/deepin-icon-theme
[22:51:32] <buildmaster> ... (24 lines total)
[22:51:44] <tyzoid> interesting
[22:51:55] <tyzoid> why'd it do it twice?
[22:52:11] <tyzoid> buildmaster: wtf watch
[22:52:15] <abaumann> indeed. bonus wtf.
[22:52:19] <tyzoid> abaumann: using buildmaster is like writing a poem
[22:52:26] <abaumann> I noticed. :-)
[22:52:37] <buildmaster> testing/bash-completion
[22:52:37] <buildmaster> testing/breeze
[22:52:37] <buildmaster> testing/oxygen
[22:52:37] <buildmaster> testing/xcursor-themes
[22:52:37] <buildmaster> community-testing/deepin-icon-theme
[22:52:37] <buildmaster> ... (24 lines total)
[22:52:37] <buildmaster> testing/bash-completion
[22:52:38] <buildmaster> testing/breeze
[22:52:38] <buildmaster> testing/oxygen
[22:52:39] <buildmaster> testing/xcursor-themes
[22:52:39] <buildmaster> community-testing/deepin-icon-theme
[22:52:39] <tyzoid> Picking words that rhyme / Obscure as they seem / Elusive as dreams / Might take some time
[22:52:40] <buildmaster> ... (24 lines total)
[22:52:49] <tyzoid> lol, it did a dup again
[22:52:51] <tyzoid> wtf
[22:53:36] <tyzoid> buildmaster: wtf linux
[22:53:52] <abaumann> I let you play.. I have to stop for today. CU.
[22:53:56] <tyzoid> see ya
[22:54:01] -!- abaumann has quit [Quit: leaving]
[22:54:12] <buildmaster> testing/ntp
[22:54:12] <buildmaster> community-testing/binwalk
[22:54:12] <buildmaster> community-testing/neofetch
[22:54:12] <buildmaster> community-testing/pm2
[22:54:12] <buildmaster> community-testing/radare2
[22:54:13] <buildmaster> ... (14 lines total)
[22:54:13] <buildmaster> testing/ntp
[22:54:14] <buildmaster> community-testing/binwalk
[22:54:14] <buildmaster> community-testing/neofetch
[22:54:15] <buildmaster> community-testing/pm2
[22:54:15] <buildmaster> community-testing/radare2
[22:54:16] <buildmaster> ... (14 lines total)
[22:54:43] <tyzoid> buildmaster: why don't you stabalize linux
[22:57:32] <tyzoid> buildmaster: why don't you stabilize firefox
[22:57:33] <buildmaster> Sry, "why-dont-you stabilize" is unavailable, until someone recodes it to look into the database.
[22:57:33] <buildmaster> Sry, "why-dont-you stabilize" is unavailable, until someone recodes it to look into the database.
[22:57:56] <tyzoid> girls / deep42thought: ^ looks like there's two buildmaster IRC listeners
[23:01:41] * buildmaster resumes sanity.
[23:01:49] <tyzoid> buildmaster: why don't you stabilize firefox
[23:01:50] <buildmaster> Sry, "why-dont-you stabilize" is unavailable, until someone recodes it to look into the database.
[23:01:50] <buildmaster> Sry, "why-dont-you stabilize" is unavailable, until someone recodes it to look into the database.
[23:41:57] -!- isacdaavid has quit [Quit: Leaving.]
[23:50:17] <buildmaster> firefox-developer-edition is broken (says buildknecht2).