IRC Logs for #crux-devel Wednesday, 2013-02-13

juegood morning00:48
juefrinnst: just saw that you are the packager of libmng, may I ask you to adopt the port?00:54
frinnstsure, no problems01:36
*** sepen has joined #crux-devel01:49
sepenjaeger: here?02:44
sepenjaeger: got some errors with new opt pre-receive hook:
sepenI think /var/log/git/hooks should be writable by opt group at least02:47
sepenjaeger: also pre-receive needs to initialize variable ERROR=0 to avoid the error: '-eq: unary operator expected'02:50
*** jue has quit IRC03:33
frinnsthe's probably not awake yet :)04:46
frinnsti noticed the same thing btw04:46
Romsteri don't thin i'm awake either :D04:59
jaegerJust woke up, I will fix it when I get to work soon, sorry05:04
jaegerok, tweaked both the scripts a bit05:55
jaegernow they will log to /var/log/git/hooks/ports/opt/<hook>.log05:55
jaegerfixed perms on the opt dir and that way we'll have proper perms on the other repo log dirs when they get created eventually05:56
jaegerOn my test git server I used gitosis so everything ran as the 'git' user instead of multiple individual users05:57
jaegerdeus_ex: testing the nvidia thing with 3.7.6 and 3.7.7 today06:22
pedjajaeger: ok.Did you come up with some cleaner/more robust solution, iirc you had some doubts about patch from FS?06:27
pedjafrinnst: how much will you be charging roelof for tech support ? :)06:33
jaegerpedja: my "doubts" are just that we'll have to update it for every new kernel version until nvidia releases a newer driver06:55
jaegerfor example, in the ticket you changed the sublevel to 6 instead of 5, but kernel 3.7.7 is out now06:55
jaegergotta go troubleshoot something, back soon06:55
pedjajaeger: I should have changed sublevel to 9 :)06:56
pedjaor 9906:56
pedjayou never know with Linus...06:57
Romsterwhat's up with the nvidia driver and kernel?07:02
sepenjaeger: we're running gitosis on crux-arm too07:14
sepenjaeger: there is still a little syntax error on post-receive07:18
sepenpython 2.7 required maybe?07:20
frinnstpedja: $200/minute07:45
frinnsthonestly i think i'd rather kill myself07:46
pedjafrinnst: there is no cluebat big enough for him, imho07:47
*** joe9 has quit IRC08:48
jaegerugh, hadn't thought of that, my python is much newer than's, probably08:58
*** jue has joined #crux-devel08:59
jaegerI ended up having to replace a laptop, took a lot longer than I expected08:59
teK_IT would be a lot easier if there were no hardware glitches...09:00
teK_or printers09:00
teK_or users/customers. :x09:00
jaegerThen they would not pay me09:00
jaegerok, python 2.5 doesn't support 'with' so I'll have to rework that hook. I'll disable it for now09:01
teK_isn't it 'easily' replaceable? I did not much python but I read that it was just faster as the alternate syntax09:03
jaegeryes, easily replaceable with some try logic, no problem09:03
jaegerI just don't want to leave it half-done if I have to go work on something again :)09:03
teK_ok, just wondering09:04
*** joe9 has joined #crux-devel09:12
jaegerpedja: it's nothing more than a workaround in my opinion but what I'll do for now is this: sed -i -e '1706s/le 5/le 99/' kernel/conftest.sh09:22
jaegerIf a newer kernel comes along that breaks it in some other way I'll investigate at that time, no way to know in advance =/09:22
frinnsthooks/pre-receive: line 37: /var/log/git/hooks/ports/opt/pre-receive.log: Permission denied09:25
jaegerfixed... logging this would be useful but it's proving to be a pain with our weird-ass git setup09:26
frinnstjue: do you have a ck4up entry for libmng? stupid sourceforge09:29
frinnsthm, the download-site on seems pretty static, should probably be able to monitor that with ck4up09:30
jaegerok, reenabled the post-receive hook09:32
jaegerI guess I should have tested it on the nvidia push but I didn't09:32
jaegerI wonder if I could run it manually09:34
juefrinnst: yeah, have on, but not very nice ->   /files/libmng-devel/[.0-9]+|@TAR@09:35
juejaeger: can push something to opt if that helps?09:36
jaegerjue: if you need to, go ahead... but I tested by running the post hook manually and it sent the mail :)09:37
jaegerthe pastebin link I just pasted is the log09:38
jaegerI'll set up logrotate for that as well09:38
jaegerif we use that :D09:38
jaegerlooks like we don't09:38
jaegerMaybe when the upgrade happens we can revisit it09:39
jaegerI got your cups-filters email as well, looked fine09:44
jaegercharming, sshfs stopped working with my 3.7.7 boot11:26
sepenjaeger: seems that same happened from python 2.5 -> 2.6, and worked by adding: 'from __future__ import with_statement'11:31
sepenfrom __future__ just rocks!!11:31
jaegerhaha, nice11:38
*** mike_k has quit IRC13:55
*** Lukc has quit IRC13:58
*** Lukc has joined #crux-devel13:59
*** Lukc has quit IRC14:04
*** Lukc has joined #crux-devel14:10
*** sepen has quit IRC18:02
*** joe9 has quit IRC18:27
*** joe9 has joined #crux-devel18:52
*** mavrick61 has quit IRC19:24
*** mavrick61 has joined #crux-devel19:26
*** mike_k has joined #crux-devel22:53
*** deus_ex has joined #crux-devel23:35

Generated by 2.11.0 by Marius Gedminas - find it at!