IRC Logs for #crux-devel Sunday, 2006-10-01

*** jaeger has quit IRC01:01
*** rantolo has joined #crux-devel03:36
*** rantolo has quit IRC04:22
*** tilman has joined #crux-devel05:37
*** sip has joined #crux-devel07:02
tilmanmmh08:57
tilmanthere's still a bunch of ports in opt that are maintained by "core"08:57
tilmanwindowmaker stuff e.g.08:57
tilmanwhat about just removing the wmaker ports?08:57
tilmanit seems like nobody cares about them :]08:57
tilmanand i think we can safely drop them (not so for opt/cvs eg :D)08:57
sipsure, cptn had a black list somewhere08:58
tilmana black list of what?08:58
sipports to remove08:58
aonyou can't possibly drop cvs :D08:58
siptilman, sorry, brain not connected again, the list was of ports to move from core to opt08:59
aonit's out of date, too :(08:59
sipanyway, I think we'd probably better not changing too much before 2.309:00
sipso we can take the chance of normalizing a lot of stuff for the next release09:00
sipwhich I hope can be out in a few months, if we start working on it soon09:01
tilmanwell09:01
tilmanremoving wmaker ports is a matter of 1 minute ;)09:01
sipheh, but dropping cvs would leave people screaming :)09:01
tilmannonono09:02
tilmani'm not saying we should drop cvs09:02
tilmanwe should _try_ to have people adopt these core/opt bastards though09:02
sipheh09:03
tilmanthough i can understand nobody wants cvs :)09:03
sipI mentioned cvs as an example. Apart from that and wmaker, do you feel there are other ports which usefulness is between the two?09:03
tilmancdrtools probably09:04
tilmanthough some people might find it more important than cvs09:04
sipaye09:04
tilman(most people, actually :D)09:04
tilmanthere's no port that depends on esound09:05
sipI mean, rather than fixing arbitrary things now, I'd prefer to get thinkg in well shape for 2.309:05
tilman"arbitrary things"?09:05
tilmani think this is a pretty specific problem09:05
sip"this" is wmaker or unmaintained ports in general?09:06
tilmanports in opt that claim to be maintained by crux-ports@crux.nu09:07
sipok. I put them almost in the same catefory of core ports maintained by core-ports@crux.nu09:08
sipI mean, apart from some easy decision (wmaker: adopt or drop) there could be some port that is really needed but nobody feels like maintining09:09
tilmans/could/there really is/ :)09:09
sipheh.09:09
sipI feel the key word of 2.3 should be 'focus'09:10
sipAs in remove cruft, keep essential things09:10
aonhttp://aon.iki.fi/cgi-bin/paste.cgi?id=325ea909:10
aonok? :D09:10
sip2nd half, see you later ;)09:11
tilmanwell09:11
tilmanaon: if you committed that, the earth would implode ;)09:11
tilmansince you haven't been given +w on core/ officially09:11
tilmanyet cvs is in opt09:11
tilmanbut maintained by core09:11
tilman;)09:12
aonyes09:12
aon:)09:12
tilmanmaybe there is a client-only version of cvs somewhere?09:12
tilmanor we could disable the server in the port, so it's easy to test any new releases09:12
tilmanwithout having to worry too much about security issues09:12
tilmani think then i'd be even willing to maintain it :D09:13
aonshame that opencvs isn't ready :)09:14
aoni read somewhere that local repos stop working without server code09:15
aonperhaps that should be tested09:15
aonnah, seems to work09:21
*** jaeger has joined #crux-devel10:44
*** sip has quit IRC12:58
*** sip has joined #crux-devel15:01
*** cptn has joined #crux-devel15:03
cptnhey15:03
siphi cptn15:03
sipI think Viper is on vacation or something15:03
cptnah15:03
cptnit's not urgent :-)15:03
sipnot sure about that, he's been marked away for some day15:04
tilmanevening15:04
sipand there's a pending request for removal (ML)15:04
jaegerif it's important to do it fast, I'll do it15:05
sipjaeger, no thanks, I'll have a look at it so I learn something useful to act as a backup for the future15:06
jaegerok15:06
sipbut remember you promised to become the leader of the PR team :)15:07
cptnheh15:07
jaegerhell no, I didn't15:07
cptnI think that was maro15:08
jaegereven worse than me15:08
sipeither him or han15:08
jaegerugh15:08
cptntilman: what's your feeling regarding the pkgmk change15:09
cptn(html download handling)15:10
tilmani think "meh" describes it well :D15:11
tilmani don't hit that problem often enough to bother me15:11
tilmanie, +/- 0 :P15:11
cptnokay :-)15:11
sipitems #6 and #18 of this page: http://gradha.sdf-eu.org/textos/klingon_programmer.en.html look appropriate15:12
cptn:-)15:13
cptnShould this replace the mantra?15:13
sipheh15:13
tilmanShip it! Ship it and let them flee like the dogs they are!15:19
tilmanhahah15:19
siptilman, I was (re)reading that when I saw your mail on pkgutils :)15:19
tilman:D15:24
sipwhich repo is to be removed from contrib?15:30
tilmanmorlenxus15:31
sipthanks15:31
sippkgutils and prt-get imported into git16:01
cptncool16:02
tilmanin separate repos?16:02
sipyes, you should be able to clone them (you're in the tools group)16:02
tilmanurl?16:03
siphttp://crux.nu/Main/ContribHowTo16:03
sipreplace contrib with the appropriate ditr16:04
sipdir16:04
sipalso please let me know if you note something weird with permissions16:04
tilmanthat wiki page doesn't exist :P16:04
siplogin16:05
tilmanthat means it's _protected_?16:05
sipyes, since the git contrib is not official nor announced yet16:06
tilmanindeed16:06
tilmansorry16:06
tilmanah16:06
tilmani suggest to NOT use the GIT_FOO_BLAH env vars16:06
tilmanusing ~/.gitconfig is a better idea usually16:06
tilman:)16:06
sipnice, now adjust the wiki page :-P16:06
tilmanok16:07
cptnmmmh, I wonder whether we should extract the kernel headers for the glibc port16:07
cptnor just download the kernel source16:07
sipgit is still a new think for me so please forgive my inaccuracies16:07
cptnit's 3560k vs 40 MB16:08
cptnOTOH we have to provide it on crux.nu...16:08
siphm, not sure if using the kernel ones will lead to problems, depending on the kernel CRUX is shipped16:09
tilmansip: np, ~/.gitconfig is also new and not documented yet i think16:10
cptnsip: 2.6.18 has a 'headers_install' target16:10
tilmanthe order in which the config is evaluated is:16:10
cptnsip: which is meant to be used as glibc / user space headers16:10
tilman~/.gitconfig, .git/config (per repo), env vars16:10
siptilman, could you please tell us something about the 'remote origin' problem? (saw the xorg thread)16:10
tilmanie, env vars override everything16:10
tilmansip: basically, never use "git push -a" (push all)16:11
tilman"git push" or "git push origin" will work16:11
sipisn't this something we can 'centrally' avoid16:11
sip?16:11
tilmanuh16:11
tilmandunno16:11
tilmanpulling pkgutils.git worked btw16:12
sipgood16:12
*** pitillo has joined #crux-devel16:12
* tilman edits pkgutils/.git/config to use his crux.nu email adress16:12
tilman\o/16:12
cptnsip: so the only question is whether I should tar them up manually, or generate them at build time16:13
sipcptn, on headers, if this is now somehow reccommended, I see no reason not to use it16:14
cptn?16:15
sipI mean, I thought the common opinion was to never use kernel headers16:16
cptnyes16:16
cptnwith 2.6.18, the 'headers_install' should generate sane headers for userspace16:16
cptnmake headers_install16:16
sipoh, you forgot the 'sane' part before :)16:17
cptnI was just wondering whether the glibc port should download 40 kernel source to run that 'make headers_install'16:17
cptn40 MB16:17
cptnor whether I should run it once, tar it up and provide them on crux.nu16:17
cptn-> ~ 4 MB16:17
tilmanthe latter sounds saner16:17
siphmm, yes16:17
cptnokay16:18
cptnI'll write a short script to generate that tarball16:18
sipshall I also import pkgsync?16:30
cptnyou should probabyl ask Jay16:32
cptnit didn't sounds like he plans to spend a lot on time on CRUX in the future16:32
tilmannicely put ;)16:32
siphmm, last I heard him he said it's was simply he had little time in this (admittedly long) period16:33
sipI think that was on irc some day ago16:33
cptnwell, he also said he's in the same boat as Jukka16:33
cptnOTOH, it doesn't really matter16:34
cptnit's easily added later, or removed again :-)16:34
cptn(the git repo)16:34
tilman*nod*16:34
sipalright16:34
tilmanfor consistency's sake, i'd suggest to convert it, too16:34
sipconsistency? damn, the tools subdirs have a dozen different layouts :)16:35
sipI was asking 'cause I suspect git-svnimport will complain having a tag dir under the trunk16:38
tilmano__O16:39
tilmanrightly so!16:39
tilmanwe probably should add the tags for pkgutils and prt-get.16:39
tilmanbut i'm too lazy right now16:39
sipyou mean in svn?16:39
tilmanno, in git16:40
tilman"git tag pkgutils-x.z.y deadbeef" etc16:40
sipoh, ok16:40
sipspeaking about that I'll miss svn easy revision numbers16:41
tilmanyou can abbreviate them16:41
cptnand the 'revision' magic for flyspray will need adjustment16:41
sipnot to talk about the timeline16:41
tilmanie, if there's only one revision that starts with "deadbeef", you can just use it16:41
sipyes, but it was cool when you got a nice revision number :)16:43
sipI got 1000. yay!16:43
tilman;D16:43
cptnjust imagine you can create hex words now!16:43
cptnwith nicely crafted changes16:43
siplol16:43
tilmanheh16:44
*** pitillo has quit IRC17:02
cptnsip: how are the permissions managed now?17:11
cptnI tried to commit to 'scripts' but got an error17:11
siper, didn't put you on the tools group17:12
siptry now17:12
cptnworked fine17:12
cptnthanks :-)17:12
sipphew!17:12
sipwe'll soon have more repos than xorg :-)17:15
jaegerhrmm... why would fc-cache fail to write to its cache?17:27
* jaeger kicks fontconfig in the nuts17:27
tilmanstrace'ing it would probably be the quickest way to find out17:27
sipalready rejmerjed ?17:27
jaegeryeah17:28
jaegermay be an FS problem, that host went down badly yesterday17:29
sipI had the same problem but after updating the new config files everything worked fine17:29
jaegermakes sense... but already done that... we'll see17:30
cptngreat, I configured pkgmk.conf to use ccache, but forgot to install it17:32
siptipical chicken and egg problem17:33
jaegerheh17:40
*** sip has quit IRC18:01
*** cptn has quit IRC18:32

Generated by irclog2html.py 2.11.0 by Marius Gedminas - find it at mg.pov.lt!