[01:24] <jfer> hi all i was wondering if there were plans to package acire for ubuntu 10.10?
[01:24] <Bachstelze> jfer: if it isn't packaged yet, it will never be
[01:25] <Bachstelze> in the official repos, at least
[01:25] <persia> Why?
[01:25] <RAOF> Well, that's not strictly true.  It could be packaged in Natty and then backported, or it could possibly go through the ARB.
[01:25] <persia> Indeed.
[01:26] <jfer> ok
[01:26] <ari-tczew> jfer: do you want to pack it?
[01:27] <Bachstelze> natty is past FF, do new packages generally get an exception?
[01:28] <Bachstelze> er no, it's past import freeze
[01:28] <Bachstelze> meh
[01:28] <Bachstelze> I need to go to bed
[01:31] <jfer> i think that jono is the maintainer of the package so i might contact him
[01:32] <persia> Ubuntu doesn't really have maintainers, but he may be able to help, certainly.
[07:11] <MTecknology> what section would a -common package belong in?
[07:12] <RAOF> What section is the base package in?
[07:14] <MTecknology> RAOF: they should be the same?
[07:14] <RAOF> It's a pretty reasonable bet.
[07:15] <RAOF> The if -common package is for a library, then it's essentially part of that library and lives in the same section.
[07:21] <MTecknology> RAOF: not a library, the main package is in httpd
[07:22] <MTecknology> is there a debug section?
[07:22] <RAOF> If it's only useful in conjuction with $BASE_PACKAGE and doesn't obviously fit into any other sections, throw it in the same section.
[07:22] <MTecknology> nvm- i see it now :)
[07:23] <MTecknology> thanks :)
[07:23] <Rhonda> Yes, there is a debug section, but how would a -common package qualitfy for debug?
[07:24] <MTecknology> I was looking at the other packages provided
[07:24] <MTecknology> the -dbg packages
[07:26] <Rhonda> ah
[07:27] <micahg> if I'm fixing a FTBFS error from a sync, do I need to use -v on upload?
[07:35] <MTecknology> I have this .logrotate file (http://dpaste.com/361716/) and I need to add a command in there; can I just add that command to the bottom?
[07:36] <MTecknology> It needs to be either   kill -USR1 `cat /var/run/nginx.pid`   or   nginx -s
[07:36] <MTecknology> heh....... I feel really dumb now
[07:37] <Rhonda> Either in the prerotate or postrotate part, but isn't that there already? :)
[07:40] <RAOF> micahg: The sync has already been processed by the archive, so you only need to have the new changes in the changes file.
[07:40] <MTecknology> Rhonda: ya, that was why i felt really dumb :P
[07:41] <micahg> RAOF: k, that's what I thought, thanks
[08:16] <dholbach> good morning
[08:18] <MTecknology> dholbach: howdy
[08:43] <MTecknology> I love being able to use the excuse.. I can't do anything, it's compiling.
[08:51] <MTecknology> Is there any way to tell dh_installchangelogs to install the changelog to a different directory? to ./usr/share/doc/nginx/changelog.gz instead of ./usr/share/doc/nginx-common/changelog.gz
[08:58] <MTecknology> I'd like to do it with two different packages and a could different things... I want things installed as nginx instead of nginx-common
[09:06] <Rhonda> MTecknology: Put nginx-common as first binary package in debian/control
[09:07] <Rhonda> And actually, man dh_installchangelogs has documentation on its options ;)
[09:08] <MTecknology> Rhonda: there's an 'nginx' package too, a dummy one though; there's two packages that I want to install stuff into the one directory for too
[09:09] <MTecknology> Rhonda: I was looking at the man page for that app, and then started looking through man debhelper, still looking
[09:11] <MTecknology> maybe -ppackage is what I need?....
[09:11] <MTecknology> no..
[09:18] <geser> good morning
[09:20] <MTecknology> geser: how ya been?
[09:24] <MTecknology> geser: maybe you can help me... you've helped me out on a lot of toughies before... :)
[09:31] <MTecknology> geser: http://dpaste.com/361974/   nginx is a dummy package that selects nginx-full; I don't really think it needs the changelog in it; I'd like the nginx-doc and nginx-common package to install things to ./usr/share/doc/nginx/ instead of what they do now..
[09:31] <MTecknology> wow... 10min to ask a question.. punishment of 03:30 and booze
[09:33] <geser> MTecknology: currently no idea, try to find other split packages and check how they do it
[09:34] <MTecknology> geser: i tried but couldn't find much that didn't have files laid out the same way which means either a) they don't care or b) they don't knoe
[09:34] <geser> and check what Debian policy allows
[09:35] <MTecknology> or c) it's not allowed..
[09:36] <MTecknology> I need to learn how to find what I want in that policy
[09:36] <geser> for the -doc packages you could have a symlink in /usr/share/doc/nginx/html (or what ever) and point it to /usr/share/doc/nginx-doc/
[09:38] <MTecknology> hm.. i remember something about symlinks.. i'll try to find where i saw that
[09:38] <MTecknology> geser: does it look kinda ok from what you can see from that little bit?
[09:39] <geser> "/usr/share/doc/package may be a symbolic link to another directory in /usr/share/doc only if the two packages both come from the same source and the first package Depends on the second."
[09:40] <geser> and a footnote mentions that you need a strict dependency (= version)
[09:40] <MTecknology> geser: where at is that?
[09:40] <geser> http://www.debian.org/doc/debian-policy/ch-docs.html 12.3 Additional documentation, 5th paragraph
[09:40] <MTecknology> thanks :D
[09:42] <geser> if you would do this symlink (if it would work in your case), your -doc package would need to depend on nginx (or the other way around, depending how you symlink)
[09:42] <geser> and if you depend on the -doc package, you don't gain anything by the package split
[09:43] <MTecknology> Maybe I'm getting a little bit pedantic in my wants..
[10:35] <MTecknology> !info svn-buildpackage
[11:01] <MTecknology> I've heard that people naturally adjust to a 25hr day; being unemployed for a long time, i'm starting to really believe it
[11:02] <MTecknology> every day, I want to stay up an extra hour
[13:59] <jderose> hello, i was wondering if someone debhelper savy could help me fix the problem i describe here - https://bugs.launchpad.net/pyskein/+bug/709247
[14:00] <jderose> basically i need to make sure setup.py is called with LANG=en_US.UTF-8, but i'm not debhelper/make savy enough to get it working :)
[15:28] <doctormo> persia: ping
[15:28] <persia> Yes?
[15:33] <doctormo> persia: that's weird, my irc client says your not here. And yet here you are.
[15:43] <cjwatson> jderose: I've commented with a suggested approach
[15:43] <jderose> cjwatson: thanks so much, i've been totally stuck on this!
[15:46] <jderose> cjwatson: quick question: how do I "make your other targets depend on debian/tmp-locale"? sorry, not very make savvy
[15:49] <cjwatson> jderose: like this: override_dh_install: debian/tmp-locale
[15:49] <jderose> cjwatson: okay, thanks again!
[15:50] <cjwatson> (sorry, override_dh_auto_install in your case, but you get the idea)
[15:51] <cjwatson> jderose: also, 'override_dh_python2:\n\tdh_python3 ...' is just weird ...
[15:51] <cjwatson> why not just 'dh $@ --with=python3' in the '%' target at the top?
[15:51] <jderose> cjwatson: agreed, i copied from some other python3 packages... how should i do it?
[15:52] <cjwatson> possibly --with=python2,python3 if you want both dh_python2 and dh_python3, but I don't know whether you do
[15:52] <cjwatson> I guess not, your control file is 3-only
[15:53] <jderose> right now pyskein is python3 only
[15:53] <cjwatson> you might also want --without=python-support
[15:53] <jderose> so if i have --with=python3, i no longer need override_dh_python2?  do i still need    override_dh_auto_install?
[15:53] <cjwatson> you don't need override_dh_python2
[15:53] <jderose> what does --without=python-support change?
[15:54] <cjwatson> turns off the default call to dh_pysupport
[15:54] <cjwatson> I'm not sure how dh_python2 was being called for you at all in the first place, since dh_python2 is only used if you say --with=python2, unless something has changed
[15:54] <jderose> is that needed because dh_pysupport is python2 only?
[15:54] <cjwatson> I believe it to be
[15:55] <cjwatson> it's being phased out anyway, eventually
[15:55] <jderose> cjwatson: my adaptation from httplib2 might not be without serious misunderstanding on my part :)
[15:55] <cjwatson> you still need override_dh_auto_install since you need to modify the call to setup.py
[15:55] <cjwatson> I wouldn't know about that ...
[15:57] <doctormo> jderose: I would, what would you need to make setup.py do?
[15:59] <jderose> doctormo: well, i don't need to do anything special except make it work with python3.  i'll admit i don't understand that rules file, but looking at other python3 packages, they all seem to have rules kinda like that
[16:01] <doctormo> jderose: I see, so your issue is with the debian dh side of things rather than distutils.
[16:03] <jderose> doctormo: right, although there is a bug in distuils i'm also https://code.launchpad.net/~pyskein/pyskein/packaging
[16:04] <jderose> i'm also working around
[18:58] <jderose> cjwatson: i'm still getting the same build failure, would you mind looking at my rules to see if you spot any problems? http://bazaar.launchpad.net/~pyskein/pyskein/packaging/view/head:/debian/rules
[19:11] <MTecknology> ok.. there was a way to see what package installed what file, right?   what was that?
[19:13] <jderose> MTecknology: dpkg -S /usr/bin/inkscape
[19:13] <ari-tczew> jderose: could you show us error message?
[19:13] <jderose> http://launchpadlibrarian.net/63030182/buildlog_ubuntu-natty-amd64.pyskein_0.7.0-0ubuntu5_FAILEDTOBUILD.txt.gz
[19:13] <MTecknology> thanks :D
[19:13] <jderose> UnicodeDecodeError: 'ascii' codec can't decode byte 0xc3 in position 71: ordinal not in range(128)
[19:14] <jderose> this is actually the result of an upstream Python bug when no locale is set, so a need to create a locale when it's built on build server, but i can't get it working
[19:15] <ari-tczew> python 3.2? wow, natty bases on 2.7
[19:15] <jderose> PySkein is python3 only ATM
[19:15] <jderose> plus, python3 is the future :)
[19:20] <bencer> hi all, i just uploaded a new version of libredis-perl to debian and filled this but to sync the package for natty, #709431. is there anything else i should do?
[19:23] <geser> bug #709431
[19:23] <micahg> bencer: best to use the requestsync tool
[19:23] <micahg> it's in ubuntu-dev-tools
[19:23] <geser> which is also in Debian
[19:24] <bencer> going to have a look
[19:28] <kklimonda> jderose: barry may be a better person to ask about it. He has reported bug on python tracker about gettext not supporting LOCPATH, it may be related.
[19:30] <jderose> kklimonda: this is bug the the PySkein author pointing me to - http://bugs.python.org/issue9561
[19:31] <jderose> barry: could your infinite Python wisdom offer any guidance on this build problem? https://bugs.launchpad.net/pyskein/+bug/709247   :)
[19:32] <micahg> is there a more proper way to make a comment like this in a patch: http://bazaar.launchpad.net/~mathieu-tl/ubuntu/natty/evolution-rss/688776.stan+debfixes/revision/22
[19:34] <kklimonda> jderose: I don't think there is much that can be done, in bug http://bugs.python.org/issue10419 (which is linked in 9561, and which seems to describe exactly your situation) there is a patch that should be applied to distutils script to fix that.
[19:35] <kklimonda> jderose: other than monkey patching build_scripts at the runtime (which is an awesome idea, the kind of mad men propose) the best bet would be to wait for the fix to get into python ;)
[19:36] <jderose> kklimonda: shouldn't it be possible to create the necessary environment in debian/rules though?  i mean, it builds fine on my system, just not build servers
[19:37] <kklimonda> jderose: http://bugs.python.org/issue8409 suggests that gettext doesn't honour LOCPATH, it may be related
[19:37] <kklimonda> jderose: the fix may be to generat locales system-wide
[19:38] <jderose> kklimonda: um, how would i do that?  :)  http://bazaar.launchpad.net/~pyskein/pyskein/packaging/view/head:/debian/rules
[19:41] <kklimonda> jderose: remove $@/en_US.UTF-8 from the localegen invocation (actually you can remove whole debian/tmp-locale target and move localedef to override_dh_auto_install
[19:41] <kklimonda> no, wait - don't remove whole $@/en_US.UTF-8
[19:41] <kklimonda> just $@/
[19:41] <kklimonda> and leave en_US.UTF-8
[19:43] <jderose> if ! localedef -i en_US -c -f UTF-8 -A /usr/share/locale/locale.alias --quiet en_US.UTF-8; then \
[19:44] <jderose> kklimonda: it should be like that? ^^^  and then do i remove LOCPATH=$(CURDIR)/debian/tmp-locale?
[19:46] <kklimonda> jderose: try that http://paste.ubuntu.com/559670/
[19:46] <kklimonda> well, you can remove the LOCPATH
[19:47] <jderose> kklimonda: thank you! trying...
[19:51] <jderose> kklimonda: so on the build servers will rules have sufficient permissions to modify system wide locale?
[19:53] <kklimonda> jderose: each build is done in a chroot (or maybe vm), so they have all rights inside it
[19:54] <jderose> kklimonda: ah, gotcha.  well, i just did a dput, maybe this got it. thanks again for the help!
[19:57] <ari-tczew> kklimonda: maybe you should consider join to MOTU?
[20:00] <kklimonda> ari-tczew: I did, now I just need some time to gather endorements.
[20:01] <ari-tczew> kklimonda: create a wiki page and I'll comment
[20:01] <kklimonda> I'd have to go through my bugs, and see who did sponsor them - which is taking me more then required. But then I've always been.. less motivated on winters ;)
[20:01] <kklimonda> ari-tczew: https://wiki.ubuntu.com/KrzysztofKlimonda/MOTUApplication
[20:03] <ari-tczew> kklimonda: from my documents mean that I've sponsored for you 1 bugfix.
[20:04] <kklimonda> ari-tczew: no idea - my documentation is lacking :)
[20:04] <ari-tczew> kklimonda: but I remember that I asked you for fix bug and you prepared a patch so quickly which is good asset for comment.
[20:04] <kklimonda> ari-tczew: ah, the g-s-d
[20:05] <ari-tczew> kklimonda: no, gwget, but right, g-s-d as well :)
[20:05] <kklimonda> right
[20:09] <ari-tczew> kklimonda: have you got more to sponsor ATM?
[20:11] <kklimonda> ari-tczew: hamster-applet for natty, and same version as sru for maverick - bug 697667 and bug 654397
[20:12] <ari-tczew> kklimonda: OK, I'll get them. Do you consider sponsoring time when you be in MOTU? ;)
[20:15] <kklimonda> ari-tczew: I don't know, I don't like to touch stuff that I don't use or at least care for as I can't be sure if I'm not breaking it. But I could sponsor stuff for things I care about, or general things like ftbfs, packaging fixes, or transitions.
[20:17] <ari-tczew> kklimonda: yea, small patches like FTBFS fixing are welcome to sponsor. January is last month when I got time for Ubuntu.
[20:20] <maco> kklimonda: you're not a motu?
[20:21] <ari-tczew> kklimonda: I had a quick research of your uploaded packages and there are some sponsors which did for you single uploads (1-2, no more). dholbach, coolbhavi, quadrispro, chrisccoulson or seb128
[20:21] <maco> hmm i feel like ogra now. thats what he said to me a month before i applied for motu
[20:21] <micahg> maco: that's generally a sign it's time for someone to apply :), happened to me too
[20:22] <ari-tczew> maco: If you feel that kklimonda should join MOTU, comment his application. ;)
[20:22] <maco> ill take a look at his uploads. i just see him doing so much i assumed he was one already
[20:23] <ari-tczew> maco: maybe you want these 2 things hamster-applet to sponsor to be sure?
[20:24] <ari-tczew> My endorsement is so strong without them, so I can leave them for someone else who want to comment kklimonda's application.
[20:24] <maco> i cant do any sponsoring til i get internet access at home
[20:24] <maco> which ...assuming i get first paycheque monday...should be about a week
[20:25] <maco> (tarballs + starbucks dont mix)
[20:25] <micahg> maco: so you need the BW equivalent of the trente?
[20:25] <maco> the what what?
[20:26] <micahg> err, bandwidth equivalent of the trenta, http://www.starbucks.com/blog/653/-trenta-means-more-refreshment
[20:26] <maco> well actually...i could ssh to my vps
[20:27] <maco> 31oz? who use 31? that is not a multiple of 4!
[20:34] <kklimonda> maco: well, I just see uploading rights more as a responsibility, and less as a privilege. But I guess it's time to grow up a little ;)
[20:35] <micahg> kklimonda: in reality it's both
[20:35] <kklimonda> hmm.. once again I got lost in my directory with various packages..
[20:35] <kklimonda> can someone describe how he or she keeps his packaging environment organized? :)
[20:36] <kklimonda> I've tried various ways but it all falls apart at some point, especially when I do some work on debian packages additionally.
[20:37] <micahg> kklimonda: I have everything in /opt/source, then under that I have a folder for each release, then under that a folder for sync, merge, and each pocket (-proposed, -security)
[20:38] <kklimonda> micahg: so you don't use bazaar for packaging?
[20:39] <micahg> kklimonda: generally not, but I also have a bzr branch under each release as well
[20:39] <micahg> s/branch/folder
[20:39] <kklimonda> mhm
[20:40] <micahg> actually, hmm, there's more, I have a bzr folder under /opt/source for stuff I manage outside the distro, I also have /opt/source/daily for the mozilla dailies :)
[20:41] <kklimonda> I like the idea with creating new folders for -proposed and -security
[20:41] <hakermania> Why revu has this sign of Christianity?
[20:43] <micahg> kklimonda: for people that always track lp:ubuntu/foo, I guess you can have those in a separate folder as well
[20:43] <Pici> hakermania: Um.  What are you referring to exactly?
[20:44] <kklimonda> hakermania: it's the same logo as MOTU have on LP, I don't think it's meant to be a cross.
[20:44] <kklimonda> at least not the christian one.
[20:44] <hakermania> Pici, kklimonda: I'm just wondering if it means something...
[20:44] <micahg> it's the ubuntu-dev badge on LP actually
[20:45]  * micahg heads out for a bit
[20:47] <kklimonda> good point :)
[21:10] <ari-tczew> kklimonda: a piece of my packaging environment ;) http://people.ubuntu.com/~ari-tczew/environment/
[21:11] <ari-tczew> I don't have that structure as micahg has, but I remember (yet) which package is for what.
[21:14] <kklimonda> so I just done something funny
[21:14] <kklimonda> I wrote find . -type f -delete instead of find . -maxdepth 1 -type f -delete
[21:14] <kklimonda> good it wasn't in some important folder..
[21:14] <maco> kklimonda: i keep source packages in ~/src/ and if its stuff-to-sponsor then in ~/src/sponsor/  -- code i write goes in ~/code/
[21:15] <kklimonda> maco: and how do you keep package in ~/src ?
[21:15] <kklimonda> packages*
[21:15] <maco> one dir per source package, with its .orig, .dsc, and .diff.gz & unpackedness in there
[21:15] <ari-tczew> it's very good topic for blog or something :)
[21:16] <ari-tczew> bdrung: ping
[21:16] <maco> kklimonda: all the versions i have of that src pkg are in that dir together. though if i also maintain it in debian, theres a debian dir in there to keep debian and ubuntu separate
[21:16] <kklimonda> maco: ah, I end up with a mess when I do that really quick - like ls ~/code/maintainance/transmission |wc -l
[21:16] <kklimonda> 100
[21:16] <kklimonda>  ;)
[21:17] <maco> occasionally i delete dirs from ~/src/ for stuff ive already uploaded
[21:17] <maco> it tends to result in reclaiming several gigs of disk space
[21:17] <kklimonda> yeah, this is really nice side effect :)
[21:19] <kklimonda> I've actually started unpacking random pieces of software (like packages I'd like to take a quick look at) in /tmp so I don't have to remember about them.
[21:20] <jderose> kklimonda: hmmm, still not working, seems like build servers didn't like creating the locale - http://launchpadlibrarian.net/63035122/buildlog_ubuntu-natty-amd64.pyskein_0.7.0-0ubuntu6_FAILEDTOBUILD.txt.gz
[21:21] <kklimonda> hmm, builders may be dropping privileges after unpacking everything.
[21:23] <kklimonda> jderose: now, there is one more thing you can try doing.
[21:23] <kklimonda> jderose: adding langpack-all | language-pack-en-base to build-depends
[21:23] <jderose> ah, okay
[21:23] <kklimonda> jderose: but I'm not sure if it will work with official Ubuntu builders
[21:24] <jderose> kklimonda: and keep rules as is?
[21:24] <kklimonda> jderose: you can remove localedef call
[21:25] <kklimonda> hmm
[21:27] <kklimonda> jderose: but it may not build when you try to upload package to official ubuntu archives.
[21:27] <kklimonda> for a reason I don't know yet
[21:27] <kklimonda> (as it builds fine on ppa builders)
[21:27] <jderose> well, getting building on ppa builders will be a good step  :)
[21:39] <jderose> kklimonda: ug, still not working - http://launchpadlibrarian.net/63041366/buildlog_ubuntu-natty-amd64.pyskein_0.7.0-0ubuntu7_FAILEDTOBUILD.txt.gz
[21:40] <jderose> kklimonda: but thanks so much for all the help!
[21:40] <kklimonda> jderose: try LC_ALL instead of LANG
[21:41] <jderose> so LC_ALL=en_US.UTF-8 ?
[21:41] <kklimonda> yes
[21:41]  * jderose dputs
[21:42] <kklimonda> jderose: configuring local pbuilder would make it much faster :)
[21:43] <jderose> by default pbuilder will have locale setup thought, wont it?  how to i make it like the ppa build servers in this way?
[21:43] <kklimonda> jderose: no, pbuilder creates an empty chroot so it wouldn't have any locales set
[21:44] <jderose> ah, okay, i'll play with that then...
[21:44] <kklimonda> (empty of packages other than essential minimum)
[21:44] <kklimonda> ari-tczew: heh, how can this folder that clean (speaking of http://people.ubuntu.com/~ari-tczew/environment/env.jpeg) ;)
[21:46] <kklimonda> cjwatson: bah, looks like python ignores LOCPATH so I can't use localedef as an alternative to langpack-all | language-pack-en-base :/
[22:04] <ari-tczew> kklimonda: as I said, in general I know which package is for. TODO file helps sort out ;)