[01:46] <oix29> hi ! someone can tell me how can I propose a new feature for the next ubuntu release ?
[01:50] <jmarsden> oix29: Create a Blueprint for it, work to refine it and gain consensus around it, attend UDS and promote it... do the work to make it happen :)
[01:56] <oix29> ok good :) thnx jmarsden
[01:57] <jmarsden> oix29: You're welcome.
[07:01] <dholbach> good morning
[07:57] <benonsoftware> Hi all
[10:27] <al-maisan> hello there, a package of mine fails to install due to a failure in the postinst script -- how can I see what these errors are (so I can fix them)?
[10:30] <jtaylor> you can put a set -x into the script to see what it executes
[10:30] <al-maisan> on the shebang line?
[10:31] <jtaylor> after it
[10:31] <al-maisan> like "#!/bin/sh -x" ?
[10:31] <al-maisan> ah, "set -x"
[10:31] <al-maisan> ok
[10:31] <al-maisan> sorry
[10:32] <al-maisan> jtaylor: thanks for the advice .. will try that.
[10:51] <Rhonda> You can even run it manually.
[10:51] <Rhonda> Like, sh -x /var/lib/dpkg/info/$package.postinst
[10:51] <Rhonda> … and configure as argument.
[10:51]  * al-maisan tries that
[10:53] <al-maisan> hmm .. I see where it's failing now, thanks jtaylor and Rhonda !
[10:53] <Rhonda> de nada :)
[11:23]  * dholbach hugs tumbleweed
[11:24] <tumbleweed> heh, np
[11:44] <jtaylor> dholbach: isn't beta freeze over? (concerning bug 842211)
[11:45] <jtaylor> the sync just fixes a ftbs no new features
[11:46] <Laney> yes indeed it is over
[12:43] <blueyed> What makes Pillow/PIL's setup.py pickup /lib/i386-linux-gnu/libz.so.1 when using Ubuntu's python, but miss it when using a self-compiled python?
[12:43] <blueyed> I have seen that Python's setup.py uses dpkg-architecture, but could not find something similar in distutils.
[14:08] <al-maisan> hmm .. what user is used to run a debian/postinst script? I thought it was root..?
[14:09] <ogra_> it is
[14:09] <al-maisan> I have statements starting with "sudo -u postgres psql .." and when I install the package I am getting prompted with "[sudo] password for postgres:"
[14:10] <ogra_> you want su
[14:10] <ogra_> not sudo
[14:10] <al-maisan> aaaaah
[14:10] <al-maisan> ok
[14:10] <al-maisan> ogra_: thank you very much indeed!
[14:10] <ogra_> :)
[14:40] <dholbach> jtaylor, Laney: you're right
[14:42] <Laney> :-)
[16:24] <tumbleweed> micahg: whoah, you only just beat me in the DMB election, well done :)
[16:24]  * tumbleweed was not expecting to do well
[16:24] <nigelb> Oh, the results are out?
[16:25] <Laney> yet to be confirmed
[16:26] <tumbleweed> right, sorry Laney :)
[16:26] <Laney> :P
[18:15] <jtaylor> I'm wondering if bug 842985 needs an ffe, its technically a new feature, but a feature which worked in natty (= regression)
[18:16] <jtaylor> brb
[18:34] <c_korn> how can I tell dh_install (via the .install file) to install a directory with a space in it?
[18:35] <c_korn> lib/mmc\ Libs usr/lib/mmediac <-- this does not work
[18:38] <cjwatson> c_korn: You can't.  http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=198507
[18:39] <cjwatson> Well, there is a workaround in there.  But personally I would just do such horrible things by hand in debian/rules.
[18:40] <c_korn> hm, ok. thanks.
[18:44] <c_korn> hm, dh_makeshlibs and dh_shlibdeps also fails :/
[18:44] <c_korn> think I need to rename the directory at the end
[19:07] <Cypher2_> hello
[19:09] <cypher2__> hello
[19:10] <cypher2__> my name is nathanel titane (Cypher2) and am currently experiencing difficulty pushing a fix to lp
[19:11] <cypher2__> anyone?
[19:12] <cypher2__> hello?
[19:13] <cypher2__> help?
[19:14] <cjwatson> I think you should give more detail including a pastebinned transcript of what you're trying to do and what the output is, and then somebody will probably look at it
[19:14] <cjwatson> (that works better than waiting for somebody to volunteer first)
[19:23] <jtaylor> does using dh-autoreconf to fix a build failure need a FFe?
[20:02] <pabelanger> Afternoon, I was looking for help triaging a backport request: https://bugs.launchpad.net/lucid-backports/+bug/638213
[20:02] <pabelanger> I'm told this is the place
[20:09] <micahg> pabelanger: looks, good, but I don't have time to comment ATM, please reset to confirmed per https://help.ubuntu.com/community/UbuntuBackports#How_to_Help
[20:10] <Laney> pabelanger: we need confirmation that the no-change backport builds, installs and runs
[20:10] <Laney> aha, says that
[20:11] <Laney> pabelanger: thanks, approved
[20:12] <jtaylor> does using dh-autoreconf to fix a build failure need a FFe?
[20:12] <pabelanger> micahg: Laney: Great, thanks for looking
[20:12] <jtaylor> allows patching configure.in instead of configure which is ackward
[20:13] <Laney> jtaylor: wouldn't think so, providing the build system is well behaved
[20:13] <Laney> make sure the patch is forwarded etc
[20:14] <jtaylor> already "applied upstream" (= upstream removed autotools and use scons)
[20:14] <Laney> fun
[20:15] <jtaylor> but we can't use that for oneiric, new features, still waiting in mentors for debian upload, so I fixed the old build
[20:15] <jtaylor> slurm is the package
[20:40] <tumbleweed> jtaylor: re pacpl, I'd say go ahead, it sounds regressiony
[20:40] <tumbleweed> and no, I wouldn't say dh_autoreconf is particularyl dangerous
[20:40]  * tumbleweed was reviewing all dh_python2 transitions before we asked for FFes for them, and finding lots of issues (should have posted those to ubuntu-devel...)
[20:42] <ajmitch> tumbleweed: regressions like FTBFS because it tried to install into /usr/local/? :)
[20:44] <tumbleweed> ajmitch: the FTBFS ones were obvious
[20:44] <tumbleweed> more common was missing byte-compilation, or byte-compiling the wrong things
[20:44] <ajmitch> I still feel like I need someone to check over what I've done to convert them, even on the simple ones
[20:45] <ajmitch> there's still much magic that goes on behind the scenes
[20:50] <spartan-11510> Hi, i'm a begginer and i need some help
[20:50] <spartan-11510> I've found a bug and i try to patch the package, but an include is make and the file doesn't exist
[20:51] <spartan-11510> The gettext.h doesn't exist
[20:51] <spartan-11510> /* Internationalization.  */
[20:52] <spartan-11510> #include "gettext.h"
[20:52] <spartan-11510> #define _(str) gettext (str)
[20:52] <spartan-11510> #define N_(str) gettext_noop (str)
[20:56] <spartan-11510> it's the hello-2.6 package
[21:01] <jtaylor> spartan-11510: the file is in gnulib/lib in that package, you ahve to pass that to the compiler with the -I flag
[21:02] <jtaylor> spartan-11510: what exactly are you trying to do? the makefile should do that for you in that package
[21:03] <spartan-11510> When you use "hello -h" he said we can use -m option but if you try to use ii, it doesn't work
[21:05] <jtaylor> spartan-11510: it doesn't offer the -m option for me
[21:06] <spartan-11510> maybe it's translation, i'm french
[21:06] <spartan-11510> it proposed me for read e-mail...
[21:11] <jtaylor> spartan-11510: I don't understand, hello is just an example application, it only prints out hello world not more
[21:12] <spartan-11510> I know, but if you look in the translation file hello-2.6/po/fr.po you can see that
[21:12] <spartan-11510> #~ msgid ""
[21:12] <spartan-11510> #~ "GNU hello, THE greeting printing program.\n"
[21:12] <spartan-11510> #~ "\n"
[21:12] <spartan-11510> #~ "Usage: %s [OPTION]\n"
[21:13] <spartan-11510> #~ "\n"
[21:13] <spartan-11510> #~ "  -h, --help          display this help and exit\n"
[21:13] <spartan-11510> #~ "  -v, --version       display version information and exit\n"
[21:13] <spartan-11510> #~ "  -t, --traditional   use traditional greeting format\n"
[21:13] <spartan-11510> #~ "  -m, --mail          print your mail\n"
[21:13] <spartan-11510> #~ "\n"
[21:13] <spartan-11510> #~ "Report bugs to bug-gnu-hello@prep.ai.mit.edu.\n"
[21:13] <spartan-11510> #~ msgstr ""
[21:13] <spartan-11510> #~ "Le logiciel Hello de GNU, LE programme de Bienvenue.\n"
[21:13] <spartan-11510> #~ "\n"
[21:13] <spartan-11510> #~ "Usage: %s [OPTION]\n"
[21:13] <spartan-11510> #~ "  -h, --help          afficher l'aide-m�moire\n"
[21:13] <spartan-11510> #~ "  -v, --version       afficher le nom et la version du logiciel\n"
[21:13] <spartan-11510> #~ "  -t, --traditional   utiliser le format traditionnel de salutations\n"
[21:13] <ari-tczew> !pastebin | spartan-11510
[21:13] <spartan-11510> #~ "  -m, --mail          afficher le courrier\n"
[21:13] <spartan-11510> #~ "\n"
[21:13] <spartan-11510> #~ "Rapporter toutes anomalies � bug-gnu-hello@gnu.ai.mit.edu.\n"
[21:14] <sladen> spartan-11510: -> http://pastebin.ubuntu.com/
[21:14] <jtaylor> interesting, that is indeed a bug and its in a lot more translations too
[21:15] <spartan-11510> http://paste.ubuntu.com/683865/
[21:15] <spartan-11510> I'm a begginer and i try to progress in packaging and help the community
[21:16] <spartan-11510> And i'm sorry for multi-line text
[21:17] <spartan-11510> What i can do just modify the translation?
[21:19] <jtaylor> spartan-11510: first you should check if it has been fixed or reported upstream, if not do so
[21:20] <spartan-11510> on launchpad? because i think it's a debian package not an ubuntu
[21:20] <jtaylor> upstream in this case is gnu
[21:21] <jtaylor> upstream= the people who wrote the software
[21:21] <jtaylor> spartan-11510: to create a patch simply fix the po files and do debuild -us -uc -S, this will create a patch in debian/patches/debian-changes-2.7-1
[21:22] <jtaylor> you can submit this patch to upstream or open a bug in debian and attach it there
[21:22] <spartan-11510> ok thank's
[21:23] <jtaylor> but note, this is not a very important application, its possible nobody will care
[21:27] <spartan-11510> debuild -us -uc -S doesn't work
[21:28] <jtaylor> please pastebin the output: debuild -us -uc -S | pastebinit
[21:32] <spartan-11510> jtaylor: result http://paste.ubuntu.com/683876/
[21:33] <jtaylor> spartan-11510: you still have a backup/swap file in the build tree from your editor, remove them before building
[21:34] <spartan-11510> all *~ file
[21:34] <jtaylor> what debuild will do is take your extracted tree and compare it to the original unmodified source tarball and make a patch from it, it cannot handle binary files that way
[21:34] <jtaylor> (unless you explicitly tell it too, but thats rarely needed)
[21:38] <spartan-11510> i haven't *~ file
[21:40] <jtaylor> hello-2.6/src/.system.h.kate-swp
[21:42] <jtaylor> spartan-11510: I just quckly check, it was already reported but ignored https://lists.gnu.org/archive/html/bug-hello/2010-12/msg00003.html
[21:42] <jtaylor> you could reply to that message telling them that the problem are the translations
[21:43] <spartan-11510> jtaylor: ok, and i could send the patch
[21:43] <jtaylor> yes
[21:48] <spartan-11510> jtaylor: thank you for your help now i know how to debug a simple package
[21:49] <jtaylor> spartan-11510: when this worked you can now test the patch by building the package completely
[21:49] <jtaylor> spartan-11510: debuild -us -uc  (without -S), this will create a .deb in the parent folder which you can install
[21:50] <jtaylor> for details on the flags see man dpkg-buildpackage
[21:57] <spartan-11510> jtaylor: thank you very much for everithing but i build it and i make dpkg --install hello_2.6-1_amd64.deb
[21:57] <spartan-11510> and i've always  the same package
[21:59] <spartan-11510> always same error
[22:00] <spartan-11510> i've make
[22:00] <spartan-11510> dpkg-buildpackage
[22:00] <spartan-11510> dpkg --install hello_2.6-1_amd64.deb
[22:03] <jtaylor> you mean the --mail is still there?
[22:05] <spartan-11510> yes
[22:08] <jtaylor> unfortunatly I'm unfamiliar with translations, maybe you have to edit a different file or regenrate something
[22:12] <spartan-11510> i will search on google i try to fix this, and enjoy you're night school tomorrow...
[22:12] <spartan-11510> Bye see you next time
[22:12] <jtaylor> bye
[22:12] <spartan-11510> And thank you