IRC Logs for #crux-devel Tuesday, 2008-07-15

*** sepen has quit IRC01:28
*** mike_k has joined #crux-devel01:33
*** rantolo1 has joined #crux-devel02:07
*** sepen has joined #crux-devel02:10
*** mike_ has joined #crux-devel03:08
*** mike_ is now known as Guest8559603:08
*** mike_k has quit IRC03:10
*** Guest85596 has quit IRC03:23
*** fetid has quit IRC03:55
*** Rotwang has joined #crux-devel04:45
*** Man0l0 has joined #crux-devel05:58
*** Man0l0 has left #crux-devel06:50
*** pitillo has quit IRC07:11
*** pitillo has joined #crux-devel07:12
*** rantolo1 has quit IRC08:26
*** rantolo1 has joined #crux-devel08:50
*** geom has quit IRC08:53
*** rantolo1 has quit IRC09:38
*** jmvr has joined #crux-devel10:25
*** cptn has joined #crux-devel11:25
*** fetid has joined #crux-devel11:32
*** Rotwang is now known as Roy`away11:35
*** Roy`away is now known as Rot`away11:36
*** sepen_ has joined #crux-devel11:37
*** sepen has quit IRC11:38
*** sepen_ has quit IRC11:38
*** sepen has joined #crux-devel11:39
*** geom has joined #crux-devel11:56
*** fetid has quit IRC12:12
*** mike_k has joined #crux-devel12:17
*** jue has joined #crux-devel12:42
*** Man0l0 has joined #crux-devel12:59
Man0l0hi tilman !!13:01
*** rehabdoll has quit IRC13:01
Rot`awayhai guyz13:01
*** Rot`away is now known as Rotwang13:03
cptnthe meeting, or the "away nicks suck ass" link?13:05
cptnlet me just get something to eat13:06
tilmanyou knew all day that we'd start at 8 o'clock!!11113:06
cptnno one told me13:07
cptnso, let's start with the bug front13:08
tilmanmy plan was to document stuff13:09
tilmancptn: like you explained later13:09
cptnsounds good to me13:09
tilmandidn't do it yet because writing docs is boring :p13:09
cptnI reported a bug against contrib/pypanel13:11
cptnsince I'm listed as maintainer13:11
cptnso far, no one has spoken up13:11
cptnI was wondering whether we should have meta accounts for each repository13:12
cptncore-ports, opt-ports, contrib-ports13:12
cptnthis would also make it easier for the users13:12
sepen#333 both opinions sounds good to me too13:12
sepensorry for the delay :)13:12
tilmancptn: mmh, why? :p13:12
cptnwell, i can also send a mail to contrib@13:14
cptnhowever if it helps to get bugs assigned...13:14
tilmanit probably doesn't :>13:15
cptnokay :-)13:15
cptnoh, do we have a flyspray bot?13:16
tilmanonly sometimes13:16
tilmanerr, it's there. i'm blind13:16
* tilman pokes spraybot 13:16
tilmanit uses some unusual prefix13:16
tilmanspraybot: 31513:16
aon%task 31513:17
spraybotFS#315 - PKGMK_ROOT not being escaped13:17
aonhi again13:17
tilmanwhy didn't grep find %task? o_O13:17
tilmanoh, because i grepped for spraybot :D13:17
aonthat might just explain it13:18
cptntilman: did you already chech that one out?13:18
*** rehabdoll has joined #crux-devel13:19
cptncan I assign it to you nevertheless?13:19
tilmani just assigned 315 to nipul13:19
tilmanpeople who have directories with spaces need to suffer imnsho13:20
tilmancptn: also, nobody found that bug in ... 8 years!13:20
sepencptn, #319 >;a=history;f=pypanel/Pkgfile sounds like prologic's port13:21
cptnsepen: it is, but prologic is not part of contrib anymore13:22
cptnhowever, it's still in contrib13:22
sepenyeah, what to do then?13:22
aonthat's almost as good as fixing a 33 year old yacc bug :)13:22
tilmansepen: ask around whether anyone wants to adopt it13:22
sepenreplace to unamintained?13:22
tilmanif nobody wants it, remove it!13:22
sepenok I'll do13:22
cptnI offered to pick it up13:22
cptnin the bug report13:22
sepenor I could pickup myself13:23
cptnboth should be fine13:24
cptnI'm using it on a daily basis, so I have no problem picking it up13:24
cptnbut if you want it, it's less work :-)13:24
cptnsince it already is in contrib13:24
sepenwell isn't my plain but I could do13:25
sepenso adopt it if you want, if not I send a mail to contrib or pickup them13:25
sepenjust your choice :)13:26
mike_kcptn: btw, what is a (technical) reason for a developer with opt commiting permissions to maintain some contrib ports?13:26
sepenprivileges and port usage?13:26
aondependencies being in contrib comes to mind13:26
aonand users ranting about conflicts in opt ports13:26
cptncan't think of another one than dependencies13:26
mike_kok. I just feel safer if something important appears in opt =)13:27
cptnheh, okay13:27
cptnI'd also like to avoid any "your taking away my port drama", since python-xlib is maintained by Romster13:27
cptnand it's a dependency for pypanel13:27
aonsame with herrie eg.13:28
cptnbut I guess I'll just ask him whether he minds giving it back13:28
* tilman flinches13:29
cptnalready asked :-)13:29
cptnnext bug:13:29
cptn%task 13013:29
spraybotFS#130 - pkgmk cannot gzip hardlinked man pages13:29
cptnit's been "new" since september 200613:29
cptncan we change its status?13:29
tilmanoh, yes13:30
Rotwangisnt it fixed?13:30
tilman"acknowledged" would be correct13:30
tilmanRotwang: nah13:30
tilmani'm not actively working on it of course13:32
tilmanif anyone happens to have a patch... :)13:32
tilmanpreferably a tiny patch that fixes the problem13:32
cptnchanged it to "researching"13:32
tilmanhuge patches make me cry13:32
aoni'll make a tiny patch and send it to you in a .doc13:33
tilmanthe remaing bugs are old an boring13:33
cptnnot all13:34
tilmanlink or it isn't true ;D13:34
cptnmplayerplug-in is unmaintained13:34
tilmanah, yes13:34
cptni.e. no one picked it up from predatorfreak13:34
cptnshould it go?13:34
tilmancptn: i think it should go; it's been almost two weeks13:34
tilmanmaintainer-wanted at crux dot nu13:35
tilmandid you actually create that alias?13:35
sepen%task 27013:35
spraybotFS#270 - udev, add a cdrom group13:35
cptnI didn't actually ask anyone to use the script :-)13:35
tilmansepen: ah, yes.13:36
sepenmm, I don't like the 'optical' name13:37
tilmani think we agreed that we liked the name 'cdrom'13:38
tilmani also think we agreed on having that stuff in udev :D13:38
aonyeah, optical seems completely idiotic13:38
sepenyeah, +1 'cdrom' isn't a silly name13:38
aoncdrom is slightly de-facto standard afaik13:38
aonat least i think i've seen it somewhere else13:38
Rotwangaon loves danny ;]13:38
sepenis not about danny13:39
aoni didn't even look at the bug13:39
aonbut thanks for telling who suggested it13:39
sepenbut this should be fixed/changed status13:39
tilmani'll merge the attached file to cdrom.rules now13:39
tilmanand see what happens13:40
cptnassigning to tilman ;-)13:40
*** geom has quit IRC13:41
tilmanthe problem is that i don't actually know that udev crap very well :>13:41
cptnthat seals it :-)13:41
sepenohh jaeger help us!!13:41
sepentilman, I think you could do the trick without problems if romster does it13:42
tilmanudev needs more love13:43
sepenI also have some minimal knonwledgements after doing vbox rules13:43
sepensee contrib/virtualbox/60-vboxdrv.rules13:43
sepenif you're agree I can test it too13:44
tilmansepen: first i'd like to understand what our cdrom.rules file does :D13:45
Man0l0I think bug #130 is solved replaceing in line 286 gzip -9 "$FILE" with gzip -9 -f "$FILE"13:46
*** geom has joined #crux-devel13:47
Man0l0sorry for mixing topics O:)13:47
sepenooh. nice13:47
sepen""If -f is not given, and when not running in the background, gzip prompts to verify whether an existing file should be overwritten.""13:48
tilmanobservation: i have quite a few bugs assigned to me13:49
tilmani seem to tend to not fix them though13:49
tilmancptn: would you like to take over 271?13:49
tilmanthe extract only thingy13:49
cptnsure, can do13:49
cptnI'll also take 29613:50
cptn%task 29613:50
spraybotFS#296 - Allow unprivileged user to run revdep13:50
*** Rotwang is now known as Rot`away13:50
tilmannext topic please!13:51
cptnaway nicks13:52
cptnis the next topic13:52
cptnanyone else dislikes them?13:52
tilmanRot`away: magic cookies for you if you stop that nick-changing annoyance :D13:52
tilmanre. "real" irc meetings13:53
cptnjust a second13:53
cptnregarding mplayerplug-in13:53
tilmannot sure "real" meetings would motivate people to attend who don't attent the "fake" meetings atm13:53
cptnshould we just drop it?13:53
tilmancptn: yes13:53
aonit can be reintroduced if someone needs it13:53
cptnat some point in time, I'd really to have this happending automagically13:54
aonfirefox-java-plugin can't be dropped, though :/13:54
cptnokay, I'll add a comment on flyspray13:54
cptntilman: true13:55
cptnthat was regarding the real meetings BTW13:55
cptnI guess we could say that we'll maybe have a more formal "release meeting"13:56
cptnbut other than that, I see no need for it either13:56
aoni also think that we get enough stuff done with these meetings too13:56
tilmanyes. feels productive :D13:57
cptngood :-)13:58
cptn"strip diffs in [notify] mails to prevent 500KB disasters (see the JDK/JRE mail) "13:59
cptntilman: what do you suggest?13:59
tilmanif (size > 8K)13:59
tilman  mail = mail[0..8K]13:59
tilmanbe brutal ;)13:59
cptnor mail = <link to gitweb>13:59
tilmanthat would work, too14:00
aona mangled diff is useless14:00
cptnI mean, we coul use gitweb links for all notify mails14:00
tilmanhaving small (complete) diffs inline is convinient14:00
aonyeah, it makes peer review easier14:00
tilmangitweb link for large mails would be good14:00
aonif it happens in the firs place ;)14:00
juedunno see a reason to send the footprint diff at all14:01
tilmanit's seldomly useful, true14:01
aonbut what if there's some user who upgrades her system by receiving a weekly digest of crux-commits@, printing them out in a webcafe and typing the diffs back in by hand? ;)14:03
cptna likely scenario indeed14:03
tilmani was mostly thinking about [notify] mails, ie crux@14:03
tilmansubscribers shouldn't have to expect laaarge mails there14:04
cptnif we cut notify mails to gitweb, and cut crux-commits@ mails to 8k14:04
cptnwould that be fine?14:04
aonyeah, i guess notifys don't need the diff at least14:04
tilmani think i'd not cut mails to crux-commits14:04
cptnso crux@ -> gitweb, crux-commit@ as is14:04
tilmancptn: i'd keep the comments from the diff in the mails to crux@14:05
tilmanso: imo just strip the .footprint diff from the crux@ mails14:05
cptnnot sure if that's easy though14:06
cptnbut I'll write that in my summary14:06
cptnand we can take it from there14:06
tilmansome perl person will be able to do it easily14:07
tilmani wouldn't attempt to do it in bash either :D14:07
cptnheh, okay14:09
cptn"improve pkgadd/pkgrm behaviour in pkgutils6 "14:09
cptnwhat did you have in mind here?14:09
cptncolors? ;-)14:09
aoni think we already discussed that -u should be ignored if the package is not installed14:09
tilmanpkgadd -u package_i_dont_have_yet14:09
aoninstead of exiting with an error14:09
tilman  "you need to drop -u"14:09
tilmanthat seems too careful14:10
cptnso you'd drop the need for -u altogether?14:10
tilmani wonder whether we could make pkgadd -u behave like pkgadd if the package isn't installed yet14:10
tilmani'd still require -u if the package will be upgraded14:10
tilmanbut if i install/upgrade a package i don't really care whether it's already installed, no?14:11
tilmani just want to make sure my system will have that new version14:11
cptnwell, I think it's somewhat inconsistent to only require it in one operation14:12
cptnbut haven't really thought about it yet14:12
*** treach has joined #crux-devel14:12
tilmanmaybe we can drop it completely14:13
juehmm, not convinced yet14:14
tilmanokay, np14:15
tilmanthe other thing i had in mind is trivial:14:15
tilmanmake 'pkgrm foo bar baz' work, ie remove more than one package at once14:16
tilmanthe same can be done with pkgadd i think14:16
sepentilman, and pkginfo -i ?14:16
tilmansepen: you're referring to that bug i guess?14:17
cptnI have no real opinion on that14:17
sepen%task 22014:17
spraybotFS#220 - pkginfo -i foo14:17
sepentilman, yeah14:17
cptnI miss neither of the two, but I mostly use prt-get14:17
cptnsurprise, surprise :-)14:17
tilmancptn: i don't hit it often, but sometimes it's annoying14:17
sepencptn, I'm refering to some Pkgfiles oto14:19
juebtw, I still think that overwriting dir-symlinks is a bug in pkgadd14:20
juetilman: you remember our discussion about that?14:20
sepencptn, /usr/ports/opt/qt3/Pkgfile:    if [ "`pkginfo -i | grep mysql`" ];then14:20
tilmanjue: can you explain the problem again please?14:20
cptnsepen: yeah, but this is ugly anyway14:21
sepengl-select too14:21
*** geom has quit IRC14:21
cptnsepen: I hope we can get rid of that at some point in time14:21
sepenhmm I suggest a builtin function inside pkgmk14:21
sepenor something14:21
jue/var/www is a sysmlink to /home/www here, everytime I install something with var/www/bla that sysmlink got overwritten14:21
sepencptn, not sure, remove my last words14:22
tilmanjue: mmh, yeah. it unlinks /var/www...14:22
juePer's pkgadd didn't do that ;-)14:22
tilmanat least that's my guess14:22
tilmanjue: can you file a bug please?14:23
tilmani'll have to inspect the libarchive code14:23
cptntilman: does pkgutils6 have some test scripts?14:23
tilmani'd love to have some more-or-less automated tests14:23
tilmanbut i don't want to stick a tar.gz in the repo :)14:24
cptnyeah, I remember you mentioned it once14:24
tilmanmaybe a separate test repo would work14:24
tilmanideas welcome14:24
cptnregarding pkgmk/pkgadd:14:24
cptnerr, pkgrm/pkgadd:14:24
cptnif it doesn't add too much code, why not14:24
tilmanit shouldn't14:24
tilmanadded to my .todo14:25
cptnthen we have the wiki as last point I guess14:28
tilmannow we could move some wiki pages14:28
cptnor that14:28
tilmanwell, feel free to explain the new FAQ structure14:29
cptnI have no idea yet :-)14:29
tilmani assume that was you who added it(?)14:29
cptnit was me14:29
cptnoh, I guess I forgot to log in?14:29
tilmandunno, didn't check the history14:29
cptnit's just that it's fairly old14:29
cptneven mentions devfs14:30
juecptn: might be a helpfull to enable pmwiki's action=rename ?14:31
cptnsounds good, don't know how it works yet though :-)14:31
sepenthe FAQ also has a broken link atm to
juecptn: add the following to local/config.php14:33
cptnokay; I'll have to download these first14:34
cptnmmmh, can't see any effect yet14:36
cptndid we maybe override the footer14:37
juego to the page you want rename and add ?action=rename to the url14:37
cptnokay, added it to the footer14:38
cptndo you guys think the FAQ should be in the Wiki as well?14:39
cptnor should that be one of the official docs?14:39
tilmanwiki sounds good i think14:40
tilmanwe can at least try it and see whether there's problems14:40
cptncan we quickly go through Main/* and decide what we want to keep there?14:40
tilmanstarting with Main/Documentation?14:41
cptnsounds good14:41
cptnHandbook should become a new top level item IMO14:41
cptnor maybe not, that's to be discussed :-)14:41
cptnbut that one should definitly stay14:42
tilmani think i don't care about that :D14:42
tilmananything below the <hr> can go to the wiki imo14:42
cptnI think all the rest could go14:42
tilmanie just leave the handbook and the release notes14:42
cptnDownload stays as is too14:42
tilmanDevelopment really needs to be cleaned up14:43
tilman"Ongoing discussions" == discussions that were ongoing 3 years ago :x14:43
tilmanerr, going on14:43
treachprovided that those issues were actually settled ;)14:44
cptnthese should go to the "History" section :-)14:44
tilmanthey weren't14:44
tilmanbut these days nobody cares about them14:44
cptnmaybe "Development" and "Community" could be merged14:44
tilmanbut the wiki page makes it look like we do14:44
aona lot of that's done, right?14:45
cptnwell, I think the "Developer Documentation" could be moved to "Documentation"14:45
cptnWebSwitch should go away14:45
cptnChrootTesting is wiki material14:46
aonglibc 2.4 :)14:46
tilmansome of the PkgutilsIdeas still sound nice, of course14:47
cptnit's more of a work in progress14:48
cptnand somewhat internal14:48
cptnI've moved a ChrootTesting to Wiki14:49
cptnand the Developer Docs to "Documentation"14:49
cptnwhat about "Community"?14:50
tilmanthat page can stay as is i think14:50
aonwell, mailing lists should be in the non-public wiki imo14:50
cptnJust wondering if we need a top level navigation item for it14:51
tilmani think so14:51
aonyeah, it doesn't seem intuitive to put it under anything else14:51
aonthere could be some better word than "community" but i can't think of one :)14:52
cptnafk for a couple of minutes14:52
*** mike_k has quit IRC14:54
sepenalso could be moved 'The team' section from to
sepenso this line "To contact the team members, please use the mailinglists or IRC channels" is redundant imho14:58
aoni don't think it's a good idea to move it to community14:59
aonsince the "community" probably is everyone else than core/opt people14:59
sepencore/opt/contrib you mean no?15:00
cptnmmmh, come to think of it if we keep the "Development" section I probably moved the dev docs too soon15:00
cptnjue: thanks for the "back to..." link :-)15:01
* sepen afk15:01
tilmanmeeting eof, right?15:12
cptnpretty much15:13
tilmangood, i should zzZz15:13
aonme too15:13
cptngood night15:14
*** Man0l0 has quit IRC15:57
*** Rot`away is now known as Rotwang17:23
*** treach has quit IRC17:28
*** jue has quit IRC17:44
*** Rotwang has quit IRC18:15
*** sepen has quit IRC18:34
*** pitillo has quit IRC18:46
*** pitillo has joined #crux-devel18:46
*** pitillo has quit IRC18:55
*** pitillo has joined #crux-devel18:57
nipuLsorry i couldn't make it to the meeting, happy spraybot was of some use :)20:59
nipuLi do like the !FS#nnn style20:59
nipuLperhaps i'll change it to that20:59
*** spraybot has quit IRC20:59
*** spraybot has joined #crux-devel21:05
*** fetid has joined #crux-devel21:05
*** spraybot has quit IRC21:06
*** spraybot has joined #crux-devel21:16
spraybotFS#315 - PKGMK_ROOT not being escaped21:16
nipuL!FS#315 assigned21:16
spraybotFS#315 - Assigned to:21:16
*** jmvr has left #crux-devel21:33
*** fetid has quit IRC22:38
*** fetid has joined #crux-devel22:41
*** geom has joined #crux-devel23:47

Generated by 2.11.0 by Marius Gedminas - find it at!