IRC Logs for #crux-devel Saturday, 2007-03-24

*** Roomster has quit IRC00:15
*** Roomster has joined #crux-devel00:56
*** Roomster is now known as Romster01:20
tilmanprologic: pong03:09
prologicaye03:09
prologicI upgraded to 2.3 fairly okay03:09
prologicupdating ports now - having problems updating subversion-perl though :/03:10
prologiclots of errors03:10
prologicthe python bindings build okay though03:10
tilmann/m03:10
*** prologic_ has joined #crux-devel04:11
*** prologic has quit IRC04:21
*** treach has joined #crux-devel05:02
*** Romster has quit IRC05:42
prologic_who has root access to crux.nu ?05:47
*** prologic_ is now known as prologic05:47
tilmanwhy are you asking?05:48
prologichave you frozen 2.2 yet ?05:48
prologicor rather emailed the ml05:48
prologicthe default branch in gitweb for contrib needs to be updated05:48
prologicthe rss feed still retrieves the 2.2 branch afaik05:48
tilmanoh, right05:49
prologicalso I'll update the ContribHowTo if you like to match git-1.5 changes05:49
prologicand a few things I've come across that'll help keep rubbish out of the tree05:49
tilmani think viper was onto that05:49
prologicalright05:49
prologicihmo it should be updated soon :)05:50
* prologic looks at the user list05:50
prologicor people like Danny will continue to commit rubbish in05:50
prologicand do silly merges05:50
*** Romster has joined #crux-devel05:51
prologichere he is :)05:51
prologicffs06:04
prologic:/06:04
prologicI've been committing a file into contrib left over by a script of mine :/06:04
prologicfuck06:04
tilman<  prologic> or people like Danny will continue to commit rubbish in06:04
tilmanPWNED06:04
tilmanprologic: did you push it yet?06:05
tilmanoh my06:06
prologicI've been pushing it all along06:10
prologicand didn't realize06:10
prologicnoone else realized it seems either06:10
prologic:/06:10
tilmanadd that fucking file to .gitignore06:13
tilman(and dont commit .gitignore)06:13
*** sepen has joined #crux-devel06:14
prologicI feel so suptid :)06:19
prologicleast I have an excuse!06:19
tilmanmaybe you should start to double-check your commits06:23
tilmanbefore you push06:24
tilmangit show sha1-hash-here06:24
tilmanwill show the commit06:24
prologic*nods*06:27
prologicI should have dbl-checked06:27
prologicI use a script to sync my repo against contrib for ports I've marked for inclusion in contrib06:27
prologicunfortunately I didn't put "git-commit.tmp" in .git/info/exclude :/06:28
tilmanyou mean .gitignore?06:29
prologicI use .git/info/exclude06:29
prologicseems to achieve the same thing06:29
tilmanmmh06:29
tilmanthen .git/info/exclude is better for your requirements06:30
prologicbut maybe .gitignore is not repo specific ?06:30
tilmansince it won't show up in 'git status'06:30
tilman(.gitignore would)06:30
prologicyeah06:30
prologicindeed06:30
tilmanshould it is ;)06:30
prologichow dare someone push in front of me :/06:32
prologiclittle bugger :P06:32
tilmanand don't push if you added merge commits to your local tree ;)06:32
tilmanlike sepen just did06:33
sepenwhat?06:33
tilmansepen: use fetch and rebase, like i wrote in the mail to crux-devel :)06:33
sepenah sorry06:33
tilmannp06:33
sepenI had remove the thunar port, it's on the xfce repo06:34
sepenmy port was obsolete xD06:34
prologicyeah I don't :)06:34
prologicI don't think you'll find a single "Merge commit" by me :)06:34
prologic*fingers crossed*06:34
prologicsepen, can you branch xfce into 2.3 ? :)06:35
sepenafter test that all depends run with it ...06:36
tilmanjue should do that, it's his repo06:36
tilmanoh06:36
tilmanyou prolly mean something else06:36
sepenjue's port is better than mine06:36
sepentilman, I'm reading ... in trouble .... someone other has modifiy the repositiory in the meantime ...06:38
tilmandid you run 'git fetch'?06:38
sepenyou had reason I need time to adjust me xD06:38
sepenno now06:38
sepenbut I'll use always in the future if I have problems06:38
tilmanoh, thought you had a problem right now06:39
sepenworking with svn I have a good practice running 'svn up' before each commit or whatever06:39
prologicI think I know the reason why I can use git-pull rather than fetch+rebase :)06:40
prologicit's ebcause I don't use the contirb repo to modify my ports with06:40
sepenoooooh shit06:40
sepenhttp://crux.nu/gitweb/?p=ports/contrib.git;a=commit;h=ba3af74ad1623258272e63ef247f88022d3bbff106:40
prologicI just sync to it06:40
sepenwhy this merge? Im only tried to do a single commit x@06:41
sepencould I rollback this?ç06:42
tilmanfirst, check whether the repo is in an okay state06:42
tilmanor whether it's really broken now (probably not)06:42
sepenfecth?06:42
tilmanno06:43
tilmanby looking at it ;)06:43
sepenoh06:43
tilmansepen: in the future: before you _push_ your changes, please run "git log" and see whether you added new "merge commits" (like the one you linked)06:44
tilmansepen: and if you did, ask someone for help06:44
tilmansepen: and don't push it yet ;)06:44
prologichmm06:45
prologicI can't push now :/06:45
sepenmmm06:45
prologic$ git push06:45
prologicerror: remote 'refs/heads/2.3' is not a strict subset of local ref 'refs/heads/2.3'. maybe you are not up-to-date and need to pull first?06:45
prologicfetch+rebase don't fix it either06:45
sepenthe same message that I had06:45
prologicsepen seems to have created a tree that mine isn't a subset of anymore :/06:45
sepensorry onemore time06:46
sepenneed time to learn how to work with git06:46
prologictilman, any ideas without rm -rf and re-cloning ?06:46
tilmani'm thinking06:46
sepenreverting and the commit?06:47
tilmanthe only other idea i have is to force the push. which is bad usually.06:47
prologicyeah06:48
tilmanprologic: cloning again is probably least painful :x06:48
prologicI forced :/06:48
sepenbut what about the wrong merge I did?06:49
prologicnothing we can do about that afaik06:49
prologicjust try not to do it again :)06:49
tilmanffs06:49
tilmanwhy the fuck did you force it when i just said that that's bad usually?06:49
sepenok I'll ask the next time06:49
sepenprologic, forced?06:50
prologiccause I saw force before you said the other :/06:50
prologicsorry06:50
prologicI didn't clobber it with anything bad06:51
prologicbut sepen's merge commit is still there06:51
sepenme: nothing to commit (working directory clean)06:51
tilmanmy head hurts06:51
tilmansepen: can you fetch and rebase now?06:51
sepenyeah Im going to06:52
sepenrebase: fatal: Needed a single revision06:52
sepeninvalid upstream06:52
sepenrebase 2.3?06:52
sepenthe last rev I did?06:53
tilmanlike i wrote:06:53
tilmangit fetch06:53
tilmangit checkout 2.306:53
tilmangit rebase origin/2.306:53
sepenok06:54
tilmandid it work?06:54
sepenhttp://crux.nu/Main/ContribHowTo06:54
sepentilman, don't appear like you wrote06:54
prologicthat page needs updating06:55
prologicfor git-1.506:55
treachheh, I'm looking for some revision system to play with - git seems to be an exellent thing to stay away from. :)06:55
sepentilman, now06:55
prologictreach, bze is not bad06:55
tilmantreach: if you're the only one pushing to it, you can hardly break it06:55
prologicbzr rather06:55
tilmanhttp://lists.crux.nu/pipermail/crux-devel/2007-March/002479.html06:56
tilmanhttp://lists.crux.nu/pipermail/crux-devel/2007-March/002493.html06:56
tilmansepen: ^06:56
sepenhttp://crux.nu/gitweb/?p=ports/contrib.git;a=shortlog;h=2.306:56
sepenseems ok no?06:56
treachI can't but help wondering if linus has the same problems as you guys have. :D06:56
tilmanheh06:57
sepenhmmm06:57
tilmansepen: yes06:57
tilmanprologic overwrote your merge commit06:57
sepensorry for the errors06:57
tilmansepen: but you might have to re-clone now06:57
sepenok but my wworking copy seems to be clean06:58
tilmannp06:58
prologicyeah sorry sepen  ;/06:58
tilmansepen: does your "git log" still list the merge commit from earlier?06:58
prologicdidn't wait for tilman's 2nd message :)06:58
sepentilman, only my last commit, not the merge commiit06:58
tilmancool06:59
sepenno need reclone06:59
tilmanwe might have been lucky06:59
sepenthat's great for me,06:59
sepenbut we need to update the wiki document06:59
tilmanviper is already working on it06:59
prologicyeah viper aparently is doing that07:00
sepenoh great! and another question, how about xml in the portdb web page?07:00
sepencan be a ticket on the flyspray?07:00
sepens/can/could ¿07:00
tilmansepen: i'll open the ticket07:01
sepenok good! Im plained to rewrite my 'portsearch' with xml support07:01
sepenor rewriting on python or pearl with libxml and libcurl07:01
sepenand about porting 'portsearch' to contrib, I should do this or wait to integrate with prtutils?07:03
tilmantreach: where's your changes for the handbook again?07:03
tilmansepen: i'd put it on hold until you have a version that uses xml07:03
tilmansepen: currently, the script is so slow and it isn't much fun to use IMHO07:03
sepenyeah according to this07:05
sepenI need to rewrite a better parser and xml could be the solution07:05
sepenthat's the actually port http://mikeux.dyndns.org:8000/crux/browser/branches/2.3/ports/portsearch/Pkgfile07:07
sepeneasy to mantain with svn07:07
treachtilman: http://hem.bredband.net/treach/Why.txt07:07
tilmansepen: https://crux.nu/bugs/?do=details&id=17507:07
sepentilman, Im going to rewrite using this temporal xml output07:09
tilmanmmh07:09
sepenbut I think i shouldn't use bash for rewrite the code07:10
tilmanperl might be fitting better07:11
sepenyeah I like regexps heheheh07:12
tilmani'd use perl bindings for libxml or so07:14
tilmanmuch easier, presumably :)07:14
sepenalso I have another script which is only a personal frontend for 'prt-get deptree' command07:15
prologicI'd use python and it's built-in elementree that it now has in 2.5 :)07:15
sepenhttp://mikeux.dyndns.org:8000/crux/browser/trunk/tools/prtree07:15
sepenit's useful to me, and I think that prt-get with integrate as option07:17
sepens/with/could07:17
prologicwhat's it do ?07:17
prologicseems to just print the number of deps that a port has07:17
prologicand -v prints prt-get's deptree of that port07:17
sepenhmm07:18
prologictilman, did you update gitweb's contrib default branch ?07:18
tilmanno07:18
prologicfor rss feeds that is07:18
tilmanno idea how to do it07:18
tilmani'd rather not touch it ;)07:18
prologicoh07:18
prologicwant me to find out ?07:18
prologichmm07:22
sepenhttp://paste.lisp.org/display/3862607:23
sependifferences of using prt-get deptree and prtree07:23
prologicyour link aboave didn't quite do that07:24
prologicand there were errors with the -v option about cat07:24
prologicand missing files :)07:24
sepenOo07:25
*** deus_ex has joined #crux-devel07:29
*** Viper_ has joined #crux-devel08:07
prologicoh goody08:09
prologicViper is here :)08:09
sepenIm plained to move my ports to crux/ports in my web, who can change it? Viper_ you?08:10
Viper_sepen: sorry i didn't get it; what do you want?08:11
sepenupdate my .httpup file08:12
sepennot now08:12
Viper_ah ok then we will have to update the portdb08:12
Viper_just sent me your new url and i will do it08:12
sepenah thanks!08:13
*** sepen has quit IRC08:35
*** sepen has joined #crux-devel08:37
*** sepen has quit IRC08:38
prologicguys, the following ports in opt need their footprints updated to match the new python-2.509:38
prologichttp://paste.lisp.org/display/3862909:38
tilmanprologic: wanna take over opt/pypanel and opt/python-xlib?09:41
tilman(Maintainer: cptn) ...09:42
prologicsure09:44
prologicshall I just maintain them in contrib ?09:44
tilmanwherever you want09:46
tilmani mean09:46
tilmanthey need to be removed from opt one way or another09:46
prologicokay09:47
prologicdoes anyone even use pypanel ?09:47
tilmandunno09:47
prologicwell I'lll maintain them for contrib then09:48
prologicyou can remove them from opt now if you like09:48
tilman\o/09:48
prologicdone09:54
clbUpdate from opt: 24 Mar 15:00 - python-xlib: removed || 24 Mar 15:00 - pypanel: removed10:16
tilmanaon: do you want to maintain opt/xemacs? ;)10:31
prologiclet's replace core/gcc with contrinb/tcc :)10:32
jjpkHow about no.10:37
prologicapparently the linux kernel can be built with it10:38
prologic;)10:38
tilmanthat guy...11:20
*** Viper_ has quit IRC11:21
*** Viper_ has joined #crux-devel11:30
prologicmorlenxus ?11:33
Viper_tilman: no error anymore on pushing something :)11:40
tilmanw00t11:40
tilmanViper_: you're going to update the wiki wrt git instructions for contrib, right?11:41
tilmanbased on my two mails to crux-devel?11:41
Viper_yes i've already done that11:41
tilmanoh, cool11:41
Viper_http://crux.nu/Main/ContribHowTo11:41
Viper_yes based on your two mails11:41
tilmanokay, i'll s/git-update-index/git add/11:44
Viper_ok :)11:44
Viper_hmm you want to use git-add when modifying a port?11:45
tilmanyes11:46
Viper_oh ok :)11:46
tilmanseems weird, but it's correct11:46
tilmanyou "add" the file to the next commit11:46
Viper_ah ok11:46
tilmanand it's "git rm -f directory"11:46
tilmanoops11:46
tilmangit rm -r directory11:46
tilmannot -f11:46
Viper_you're right11:47
Viper_didn't check that :)11:47
Viper_i am not quite sure whether this git-rebase stuff is correct11:48
tilmannice work though11:48
Viper_thanks11:48
Viper_could you check this, too?11:48
tilmanyes11:49
tilmanlooks good ot me11:51
Viper_ok11:51
Viper_i've also changed the developement site11:52
Viper_removed some stuff11:52
aontilman: hmm14:01
aontilman: perhaps i could..14:01
tilmandon't do it if you don't want to14:02
aonwell, i sort of need it anyway14:04
tilmancool14:05
tilmanyou need the editor or the sink?14:05
aonwhat do you use to find out which parts of xorg are required? ldd?14:05
aon:D14:05
aonboth14:05
tilmanldd is a good start, yeah14:06
*** Viper_ has quit IRC14:26
*** Viper_ has joined #crux-devel14:59
*** tilman has quit IRC15:16
clbUpdate from opt: 24 Mar 19:47 - rdiff-backup: updated footprint because of new python version; bumped release15:17
*** tilman has joined #crux-devel16:42
*** Roomster has joined #crux-devel18:06
prologictilman, ping18:18
tilmanyes18:18
prologicI think you should mention on the ContribHowTo page something about .git/config18:18
prologicfor handling branches properly18:18
prologicI ran into this with git-1.518:19
*** Romster has quit IRC18:19
prologicI added: push = refs/heads/2.3:refs/heads/2.3 to [remote "origin"]18:19
prologicand a new section18:19
prologic[branch "2.3"]18:19
prologic   remote = origin18:19
prologic   merge = refs/heads/2.318:19
prologicotherwise, git pull and git push are a pain18:20
prologicas long as you tell git what to fetch and merge, git pull works fine and does not create merge commits18:20
tilmanblah blah blah18:20
tilmanthe howto contains:18:21
tilmangit branch 2.3 origin/2.318:21
tilmanwhich is perfect18:21
tilman:)18:21
prologic*nods*18:21
prologicwhich is fine18:21
prologicbut still not quite complete :)18:21
prologicgit branch ... ... doesn't add Push ro [remote "origin"] or a new section for the branch18:22
tilmanwell18:23
tilmanit's a bit late for that now18:23
tilmanmaybe i'll think about it tomorrow18:23
prologicI shall test18:23
prologicI think the only reason it works18:23
prologicis because you tell users to fetch then rebase18:23
prologicwhich from reading should work18:24
prologicnevermind what I said then :)18:24
prologicguess you prefer fetch+rebase so people don't fuck it up so much huh ?18:24
tilman'pull' will always merge if necessary afaik18:25
tilmanno matter how you configure it18:25
tilmanrebase avoids the merge18:25
tilmanand shuffles commits around a bit instead18:26
tilmanor somehing like taht18:26
prologicyeah I know I get what you're saying18:26
prologicI sure would like to know how Danny causes merge commits18:26
prologicand how sepan did18:26
prologiccuase I can't work it out :)18:26
prologicnot that I'd want to do it18:26
prologicif I find any problems, can I update that page ?18:27
prologicI won't change how you want things done :)18:27
prologicI'll fully test what you and viper have got there18:28
tilmansure you can18:29
tilmanthe wiki page is based on my e-mails18:30
tilmanwhich i wrote as i tried18:30
prologic*nods*18:30
tilmanso i really think the way i described will work :)18:30
prologiccourse course :)18:30
prologicjust git-1.5 is rather strange and confusing at times18:30
prologicbut I think I finally have a full grasp of it18:30
prologicbtw, need your help (when you're free) with subversion-perl18:30
prologicit just will not compile :/18:30
prologicsubversion-python will and has been updated18:31
tilmani needed to convert an svn repo to git a few days ago18:33
tilmanbut the api of subversion-ruby was so strange and undocumented that i ended up using the CLI18:33
tilman;)18:33
prologicdid you use git-svn or git-svnimport ?18:39
prologicall my project repos were in svn, all git now18:39
tilmanneither, wrote my own script18:40
tilmannever tried git-svn though18:40
tilmanonly git-svnimport18:40
tilmangit-svn doesn't seem to import anything ;)18:40
prologicoh it does18:41
prologicwell it can anyway18:41
prologicit's used to track a remote svn repo18:41
tilman       git-svn - Bidirectional operation between a single Subversion branch18:41
tilman       and git18:41
tilmandoesn't sound like an importer18:42
prologicie: git svn url18:42
tilman:P18:42
prologicgit svn fetch18:42
prologicetc18:42
prologicthen add it to your local git repo via the normal means18:42
prologicit doesn't explitetly import18:42
prologicbut I used it to combine the crux iso with jaegar's svn livecd iso18:42
prologicand created a merged and branched version of the iso repo18:42
prologichttp://paste.lisp.org/display/3865118:43
prologicgit clone git://git.shortcircuit.net.au/crux-iso.git18:44
tilmanshouldn't you add --disable-neon-version-check to subversion-py too?18:47
prologicyeah probably18:48
prologicbut it compiles without that so *shrugs*18:48
tilmanyeah, but it gives an annoying warning18:48
prologicok18:48
prologicI'll merge subversion build changes with subversion-python/perl18:48
prologicbut the perl one still complains badly :/18:48
tilmani'll have a look in a few18:49
prologicta18:49
tilmanstill updating svn itself18:50
tilmanprologic: grep PYTHON subversion-perl/Pkgfile ;)18:53
prologicyeah yeah :)18:54
prologicI will fix that18:54
prologicjust superflueous crap in it18:54
prologicfixing subversion-python port atm18:54
tilmanalso, are you sure this LDFLAGS crap is needed?18:56
prologicno I'm not :)18:57
prologicI'll test that without18:57
prologicgimme min18:57
tilmanthe perl bindings seem to be broken18:58
tilmaneven if you follow the instructions18:58
prologic*nods*18:58
prologicthat's what I was trying to tell #svn18:58
prologicbut they didn't lsiten ;/18:59
prologiclisten*18:59
prologicthey are just swig bindings, however they are _broken_ ihmo18:59
tilmanyour port doesn't follow them though ;)18:59
prologicthe python swig bindings appear to be fine18:59
tilmanmake swig-pl18:59
prologichey it was based off someone else's Packaging :)18:59
prologicgimme a break :)18:59
tilmanmake DESTDIR=.. install-swig-pl18:59
tilmanshould work18:59
prologicbut doesn't right ?18:59
tilmannope19:00
tilmandoesn't uild19:00
prologicshame19:00
prologick trying subversino-python again with some changes19:00
tilmandoesn't build with the alternative rules either19:03
prologicreckon I/we should be mailing the mailing list19:03
prologicletting them know their swig bindings for perl are broken ? :)19:03
prologicI mean really broken :)19:03
*** Romster has joined #crux-devel19:05
tilmanwell, we don't use a known-good swig version19:05
prologictaht is true19:05
prologiccould try using the known good version of swig19:05
prologicI haven't tried that yet19:05
prologicif it compiles, the port will be a PITA though19:06
prologichowever I think I read somewhere that you can download swig and put it in the src build dir19:06
prologicand subversion's makefiles will use that or something - not entirely sure19:06
tilmandid you commit the svn-py changes?19:27
prologicnot yet19:28
prologicstill mucking around with it19:28
prologicoptimizing it a bit more ;)19:28
prologichence hte 7 compiles so far19:28
*** Roomster has quit IRC19:29
*** Viper__ has joined #crux-devel20:22
*** Viper_ has quit IRC20:24
*** Viper__ has quit IRC20:43
*** Romster has quit IRC21:33
*** Romster has joined #crux-devel21:34
*** treach has quit IRC21:37

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