=== zorael_ is now known as Zorael^nb === asac_ is now known as asac === tuantub_ is now known as tuantub === fta_ is now known as fta [00:33] is the latest time a package can be accepted for a release of ubuntu, just before the feature freeze? [00:36] trothigar, unless someone fills out the assorted paperworks, bribes, etc [00:37] directhex, ty === blueyed_ is now known as blueyed [01:39] neversfelde, pong [01:45] why would a patch fail to apply on a builder (papaya) but not for me? [01:47] depends, really - cruft in the maintainer's build directory, etc. Which build? [01:47] dtchen, it was on papaya, here's the log http://launchpadlibrarian.net/29433943/buildlog_ubuntu-karmic-i386.skia_0.0.1~svn20090723r287-1_FAILEDTOBUILD.txt.gz [01:49] sorry the build was 1130470 [01:50] have you verified that the patch actually applies in a pristine {s,}chroot? [01:51] I did build it with pbuilder just a few minutes ago [01:52] can you post or link to your source package? [01:53] dtchen, , thank you for help btw; http://bazaar.launchpad.net/~skia-team/skia/trunk/files/head%3A/debian/ [01:58] (sec, shop is closing) [02:18] binarymutant: maybe just buildd weirdness; it sbuilds ok here, too [02:18] I'll try retry building to the ppa, thanks dtchen :) [02:19] dtchen, is this expected, and if so, why? https://bugs.launchpad.net/ubuntu/+source/pulseaudio/+bug/404370 [02:19] Ubuntu bug 404370 in pulseaudio "Second or third logged in user cannot play sounds -- gnome volume control shows "null output"" [Undecided,New] === zehrique is now known as zehrique-off [07:14] are there some docs on how to get software into Ubuntu's package manager? I'm thinking of writing a small app that would be useful [07:17] oh nevermind, found the info in the urls in the topic, thanks anyway :) [09:51] gaspa: I have opened bug for round 3: https://bugs.launchpad.net/ubuntu/+source/lablgtk2/+bug/404470 [09:51] Ubuntu bug 404470 in pcre-ocaml "[3.11.1 transition][round 3/6] Please synchronize following packages from Debian sid in Karmic" [Undecided,New] [09:54] dmentre: round 2 is now complete? I've seen there were some additional rebuilds necessary as round 2 started a little bit to early [09:55] geser: it is complete on amd64. Which packages needed a rebuilt? [09:57] geser: http://bentobako.org/ubuntu-ocaml-status/transition_monitor/ocaml_transition_monitor.html === yofel_ is now known as yofel [09:58] dmentre: the ones on i386 it looks like [10:00] geser: that's strange. For example camlidl seems to have coorectly been built: https://launchpad.net/ubuntu/+source/camlidl/1.05-12build1/+build/1130989 [10:01] dmentre: it build successfully but with ocaml 3.11.0 and not ocaml 3.11.1 [10:03] geser: ouch! you're right. :-( [10:04] geser: was this because the ocaml derived binary packages were not available on build machines? [10:05] is there any way to check that? [10:05] you can look at the build log and see which pacakges were installed [10:05] or look at the depends of the resulting binary I guess [10:06] Laney: Hmm. So there is no way to check that a depenency is available before doing the build. :-/ [10:06] not without modifying the source package [10:06] well, ... [10:07] you can check if it's published before uploading it [10:07] this would be easier if we had dw :) [10:07] anyway, must dash [10:09] Laney: how can I check it is published? [10:11] either you look into the Packages files or you click your way through LP till you get to https://edge.launchpad.net/ubuntu/karmic/i386/ocaml [10:12] which was published at the same time as the rebuild of camlzip: https://edge.launchpad.net/ubuntu/karmic/i386/libzip-ocaml [10:15] dmentre: as ocaml is now published on all archs, I'll re-do the missing rebuilds [10:16] geser: thanks a lot! [10:17] geser: I hope nobody will start round 3 before round 2 is finished on i386 [10:17] Sorry for the mess. :-( [10:18] camlbz2 from round 3 is rebuild already so it will need a 2nd rebuild but that's only one package currently [10:21] geser: camlbz2 failed to build on i386 (Depends: ocaml-nox-3.11.0 but it is not installable). Will the build robot restart the built? [10:21] dmentre: only on manual request [10:21] gaspa: ok [10:22] dmentre: it can be given-back once findlib got rebuild [10:23] findlib and facile need a core-dev to do the rebuild for i386 [10:24] geser: how to "find a core-dev"? :-) [10:24] open a new bug and subscribe ubuntu-main-sponsors? [10:25] dmentre: yes (and provide a debdiff). can do it when I'm done with the other rebuilds [10:26] I'll do it. Let me know for the other rebuilds. What is a debdiff? (sorry for being totally new to all that) [10:27] dmentre: https://wiki.ubuntu.com/PackagingGuide/Recipes/Debdiff [10:28] * dmentre reading debdiff [10:31] geser: sorry, I don't understand which should I provide a debdiff, of what? We only want to have a package rebuild. The source package is the same, there are no modifications [10:32] dmentre: the version changes, so debian/changelog needs a new entry [10:33] dmentre: http://launchpadlibrarian.net/29471934/camlidl_1.05-12_1.05-12build1.diff.gz are the "changes" to get camlidl rebuild [10:33] geser: last time james_w done that, he only asked me for a sentence that were put in the rebuild logs [10:34] geser: yes, exactly that [10:34] s/were/was/ [10:35] it depends on your sponsor [10:35] it might be easier to find a sponsor when you provide a ready debdiff which they need simply to apply and upload [10:41] geser: ok [10:41] geser: I'm planning that work [10:42] I first start by opening the bug [10:42] dmentre: yes [10:43] dmentre: done the i386 rebuilds in universe === Quintasan_ is now known as Quintasan [10:49] geser: thanks for the help and rebuilds! [10:56] geser: I redid some of those yesterday evening... [10:56] i've not had time to mail both of you, sorry [10:56] gaspa: I noticed [10:56] good [10:57] gaspa: I did the ones now listed on dmentre's page [10:57] Laney: we have dw [10:57] only automatically though as far as I know [10:57] geser: how about findlib? [10:57] I'll do that one again now [10:58] gaspa: still in main [10:58] ehm.. [10:58] james_w: how about findlib? [10:58] :P [10:59] james_w: and facile too please [10:59] ouch [11:00] it's a good idea to leave lag time so that the binaries can be published on all architectures [11:09] james_w: lag time: you mean, wait some time between rounds? [11:09] yeah, where a round depends on the previous [11:10] so there should be a few hours between ocaml entering and the rebuilds happening [11:11] for a smooth process you need new ocaml -> built on all architectures -> published on all architectures (~1:30) -> rebuilds [11:11] so a round a day should work nicely :-) [11:11] james_w: ok. I'll do that [11:12] james_w: so we should wait for tomorrow for round 3 ? [11:13] if it depends on the previous round binaries being available [11:13] you can do it quicker if you like, it just takes a little care in timing [11:15] james_w: ocaml_transition_monitor builds the status table from Packages.gz files. So if this is ok in those files, we can start the next round. Am I right? [11:15] yep [11:15] ack [11:15] & cool [11:15] :-) [11:16] the buildds use a mirror that is generally faster than the one you are probably watching, so there should be no issue with that [11:16] rebuilds uploaded [11:17] james_w: I'm using http://ftp.ubuntu.com/ubuntu to get Packages files [11:17] yeah, that's actually a mirror, so can have a lag over the real archive [11:18] and I believe the buildds use the real archive [11:24] I have updated http://bentobako.org/ubuntu-ocaml-status/transition_monitor/ocaml_transition_monitor.html so that the check is done each 6 hours [11:24] instead of eahc 24 hours [11:33] Are there any good documentation on how to create a .deb from python libraries/applications? I just came across debianpackagemanager - is this used by python package maintainers? [11:34] https://wiki.ubuntu.com/PackagingGuide/Python [11:37] thanks [14:31] is something wrong with: fiordland.ubuntu.com ? [14:32] can't I set requestsync to send via my local SMTP service ? [14:33] oh,I can [14:34] cool ! [15:32] gaspa, geser: round 2 is finished! :-) Thank you for triggering recompilation of missing i386 packages (and james_w of course!). Round 3 can now start (cf. LP bug 404470). Let me know if I need to do something [15:32] Launchpad bug 404470 in pcre-ocaml "[3.11.1 transition][round 3/6] Please synchronize following packages from Debian sid in Karmic" [Undecided,New] https://launchpad.net/bugs/404470 [15:36] gaspa: the reupload of camlbz2 wasn't necessary. as camlbz2 FTBFS on i386 a give-back would be enough [15:51] hi, how do I trigger a rebuild for a failed package? [15:51] simon-o: bump the version number in the changelog and re-upload I think [15:52] unless it was on a PPA, in which case there's a rebuild link [15:52] asking for a give-back is enough [15:52] of in case of PPA: use the "retry" link as jacob said [15:53] ok, whom do I ask? It's this build: https://edge.launchpad.net/ubuntu/+source/advi/1.6.0-14/+build/1125569 [15:53] the "retry" links also exists in Ubuntu archive (if you have upload rights for that package) [15:53] ah, it's part of the ocaml transition [15:54] geser: Yes, kind of. [15:54] It was waiting for ocaml-base-nox on powerpc [16:01] simon-o: given-back [16:02] geser: thanks, that's nice :) === cprov is now known as cprov-afk [16:05] it failed again, this time with another dependency error. I think I'll wait some days until all dependencies are built. [16:07] simon-o: lablgtk2 is in round 3 and camlimages in round 4 of the ocaml transition. you can monitor it on http://bentobako.org/ubuntu-ocaml-status/transition_monitor/ocaml_transition_monitor.html [16:10] geser: ah, ok. sorry [16:11] geser: findlib on ia64 hasn't built yet, so all round three have to wait him, I guess... [16:12] yeah, better wait before we have to redo part of it [16:13] although the packages from round 3 should either build or FTBFS now so we can give-back them later, but it's easier to wait a few more till all catched up [16:14] geser: You might be able to get NCommander to rescore stuff on the slow archs if it would help move stuff along. [16:18] ScottK: it's not that important, we have still plenty of time to finish the ocaml transition [16:18] OK [16:19] gaspa: the first rebuild of findlib on ia64 was already with ocaml 3.11.1 so need to wait here on the second rebuild, the first will also do [16:20] dmentre: would it be possible to include the other archs in your ocaml transition monitor too? so we have a better overview [16:21] hi , i followed the steps as in ubuntu wiki , have my keys if i do gpg --list-secret-keys , debuild -S -sa , says secret key not available [16:22] hemanth: compare your key-uid with the name in your changelog entry (incl. any comments in your uid) [16:23] geser: checked [16:23] i did gpg --recv also [16:24] can you please paste bin your key-uid and your changelog entry [16:24] yes [16:26] geser: fixed it , thanks :) [16:27] i had added a comment, will making the key and had missed it in the changelog [16:27] like it was (comment) [16:33] i did a dput to my ppa , it said Successfully uploaded packages. [16:33] Not running dinstall. I'm not able to find the file in my lp , plz help me [16:48] all of a sudden bash: debuild -S -sa: command not found [16:54] hemanth, debuild output log please [16:55] coolbhavi: its working now ! [16:55] hemanth, good [16:55] coolbhavi: if i did whereis debuild [16:55] it used to debuild : [16:56] now its giving the path [16:56] hmm [16:56] hemanth, okay [16:57] coolbhavi: package accepted :) [16:59] coolbhavi: Estimated build start: in 3 hours !! [16:59] hemanth, yes if its building other sources it ll take time [17:00] ok [17:02] howdy! I'm thinking of adding apport hooks to a package, but the developer instructions on the ubuntu wiki seem rather sketchy. could anyone help me? [17:03] haven't read the wiki in a bit, so i don't know if /usr/share/doc/apport/package-hooks.txt.gz would be any less sketchy === joaopinto is now known as joaopinto_brb [17:30] if there are many packages in the control file and packagename.install files for each for them as well as a install file without prefix. is the content of the install file without prefix being installed to every package? === joaopinto_brb is now known as joaopinto === foursixnine is now known as santiago-ve === santiago-ve is now known as Guest43074 === korn_ is now known as c_korn [17:53] I try to build the packages of avidemux. the plugins require the avidemux libraries for compilation. they are in debian/tmp/usr/lib. how can I add this directory to the search path? I currently get this error: /usr/bin/ld: cannot find -lADM_core [17:58] python -c 'import apport.hookutils; help(apport.hookutils)' [17:58] this seems to be the answer for available apport hooks [18:29] I swear, SourceForge bugmail is incomprehensible [18:33] c_korn: is this during compiling? [18:35] geser: yes. but -l looks for the static .a files, doesn't it? but I think those files are not in this debian/tmp/usr/lib directory. but when I have a look at the compilation instructions they should [18:36] c_korn: the upstream source doesn't know anything about debian packaging [18:36] here are the instructions: http://www.avidemux.org/admWiki/index.php?title=Install_2.5#Example_Ubuntu_Compilation [18:36] -l looks for the .so file [18:37] oh [18:37] (if you don't do a static build of course) [18:38] looking in man ld I thought it looked for static libs: http://pastebin.com/d21263aae [18:40] I could also first build avidemux without plugins and then make it build-depend on avidemux (>= 2.5) and build the plugins with the libs from /usr/lib [18:42] c_korn: it looks like those passage is wrong (at least on Linux). for a static linking --static is needed. [18:43] c_korn: please try to avoid circular build-dependencies. they are a pain === ejat is now known as myfenris [18:44] c_korn: try to figure out in which directory the build libraries are (in the avidemux source) and specify that dir in -DAVIDEMUX_LIB_DIR === e-jat is now known as Guest46514 === myfenris is now known as e-jat [18:48] geser: currently I have this code in the install target (I know this is wrong but I just wanted it to work ;-) http://pastebin.com/d4ec7918f [18:49] it is just under the installation code of avidemux: cd build && $(MAKE) install DESTDIR=$(CURDIR)/debian/tmp [18:50] that might work too [18:51] I'm not sure if LD_LIBRARY_PATH is needed, as it's only used to look up libraries for run-time (IIRC) [18:51] the linker uses the option -L to search for libraries to link (which is here probably the value for AVIDEMUX_LIB_DIR) [18:53] hm, yes but I still wonder that the error is: /usr/bin/ld: cannot find -lADM_core [18:54] * hyperair curses. why is *everything* attempting to use oss now? [18:56] * e-jat pokes hyperair [18:56] ? [18:58] which "oss" is this? [18:58] if it's the "sound" oss, then it's likely a config issue [18:59] sound oss. [18:59] what configuration could i possibly have messed up that causes flash to use oss? [18:59] *and* fail miserably [18:59] and then there's mplayer, which defaulted to oss for some strange reason [18:59] i had to manually override it in my config [19:01] oss is usually a fallback; what is hogging the alsa device(s)? [19:01] pulseaudio. [19:01] that's the only thing there is [19:02] do you have a custom /etc/asound.conf or ~/.asoundrc ? [19:02] no asound.conf, but my .asoundrc contains two extra lines [19:02] pcm.pulse { type pulse } [19:02] and ctl.pulse { type pulse } [19:02] i used it for skype back in intrepid, but then something changed in pulseaudio or the alsa plugins that miserably broke bloody skype again [19:03] you shouldn't need that at all given /usr/share/alsa/pulse* [19:03] hmm? [19:03] also, do paplay and speaker-test work? [19:03] pulse is set to be the default route [19:03] well i'll delete it then [19:03] paplay works [19:03] banshee works [19:04] and the volume control applet shows banshee is accessing pulseaudio [19:04] flash doesn't [19:04] mplayer attempts to screw around with oss, but if passed -ao pulse or -ao alsa, it works. [19:04] grep oss /etc/mplayer/mplayer.conf [19:05] #ao=oss [19:05] the config says that the default is oss. [19:05] presumably [19:05] but it's commented out so it shouldnt make a difference anyway. [19:05] and flash is a bloody temperamental beast which works sometimes and doesn't other times [19:06] i'm pretty certain that i had to create ~/.mplayer/config for that reason [19:06] Flash 10? 32-bit (and if so, using nspluginwrapper) or 64-bit? [19:07] the one that comes with flashplugin-installer [19:07] on 64 [19:07] yeah, that's the nspluginwrapper + 32-bit [19:07] are you using karmic's stock pulseaudio or the one from luke's ppa? [19:08] can you reproduce the symptom if you use the native 64-bit Flash alpha from labs.adobe.com? if you can't, it's probably ia32-libs/libasound32-plugins (again) [19:08] stock. [19:09] hmm okay, this time flash's not using oss. it's using alsa [19:09] but for some reason it didn't get routed into pulseaudio [19:10] i'll go try the 64-bit flash then [19:11] (though it went berserk the last time i tried it and refused to load anything without crashing) [19:11] (remember to purge flashplugin-installer, too) [19:11] yeah i'm doing that now [19:12] by the way, ekiga used to work in intrepid, but underruns like nobody's business now too [19:12] the same as skype. [19:12] right, known [19:12] i'll be rolling a new git snapshot of pulseaudio for testing this weekend [19:12] hmm cool [19:15] Hello, does anyone know an ld option equivalent to -Wl,--as-needed that would do the same job for compiling shared libs ? [19:17] ie. avoid linking with un-necessary libraries [19:19] hmm, nevermind, it seems -Wl,--as-needed is the correct flag for this too [19:22] I still think that -l only looks for static .a libs. [19:23] and there is no libADM_core.a but a libADM_core.so [19:23] and it is in debian/tmp/usr/lib where I pointed to === Guest43074 is now known as santiago-ve [19:24] -l looks for dynamic libs. [19:25] unless you pass -static [19:25] hm, then I don't understand why it does not find the library. [19:26] maybe really a ldconfig call is missing. [19:26] but this requires root privileges, doesn't it? [19:35] dtchen: 64-bit works with pulse === DktrKranz2 is now known as DktrKranz [19:45] hm, maybe I will put the libraries directory in an own tarball and compile the plugins alone. I can then depend on avidemux-2.5 and there is no circular dependency. I will add the plugins in the Suggests field. or would depends be better? === highvolt1ge is now known as highvoltage [20:55] nhandler: Are you around? [20:57] nhandler: If you have a moment, I'd appreciate it if you'd have a look at libio-compress-perl-2.020 in pkg-perl svn and make it build. [22:00] ScottK: around ? === _ruben_ is now known as _ruben [22:33] ScottK: Still need me to look at it? [22:35] nhandler: would you mind unsubscribing motu-release from bug 392426 ? [22:35] Launchpad bug 392426 in rednotebook "Please sync 0.8.1-1 from Debian" [Wishlist,Confirmed] https://launchpad.net/bugs/392426 [22:36] stefanlsd_: seems like there's a patch for Firefox 3.5 compatibility in Gears, see the last post at http://groups.google.com/group/gears-users/browse_thread/thread/1585bcd7531c0696 [22:36] sebner: Done [22:37] nhandler: thanks :) === snearch_ is now known as snearch [22:52] if anyone has some free time, mind checking out gfire on revu? it's been sitting around for a while and i'd like to get it finished up once and for all :) http://revu.ubuntuwire.com/p/gfire [22:56] jacob: The maintainer should now be the Ubuntu Developers not MOTU, and Standards-Version 3.8.2 [22:59] nhandler: is the maintainer email still ubuntu-motu@? [22:59] ubuntu-devel-discuss [22:59] I think update-maintainer has been updated to do the right thing now [23:00] allrighty, reuploading in a moment after a couple of other changes [23:00] are there any widespread problems with mozilla plugins these days? [23:00] NEWS file seems a bit useless, no real use in debian/docs [23:00] jacob: See https://lists.ubuntu.com/archives/ubuntu-devel/2009-May/028213.html for details about it [23:01] jacob: http://gfire.site40.net/ appears to redirect to http://gfireproject.org/ [23:02] probably better to change it in copyright and control [23:02] nellery: yeah, they recently changed their site. updated, along with a typo in control [23:03] should this have a versioned depends on debhelper 7? I vaguely remember someone saying 6 is enough [23:04] though lintian is complaining about it not being 7, so I guess I'll just leave it. [23:04] jacob: if no dh7 features are used dh6 is really enough [23:05] "W: gfire source: package-lacks-versioned-build-depends-on-debhelper 7" << ignore? [23:07] argh, you are using cdbs -.- [23:07] jacob: well, leave it then [23:08] jacob: that error was because you didn't change compat [23:08] use lintian -i to get more info on the errors [23:09] nellery: makes sense, thanks [23:10] jacob: and ChangeLog has another copyright holder and some differing years [23:10] ScottK: Attribute the changelog entry to yourself. We don't tend to attribute it to pkg-perl. You also might want to look into updating the debian/rules file to the newer format the team is using (dh-make-perl --refresh should do it) === Znova_ is now known as Snova_ [23:13] nellery: ah, didn't see the copyright in the changelog before. [23:16] uploaded a new revision hopefully addressing all of this [23:17] jacob: ok. Should be ready soon enough. [23:18] nellery: thanks for the help. :) [23:18] jacob: no problem. [23:18] (also thanks nhandler & sebner) [23:38] jacob: it's not necessary to have the copyright names twice, just change the years in the first entry [23:38] for example 2000-2001, 2006-2009 for Beat Wolf === vorian is now known as OldSchool [23:40] nellery: ok [23:42] & re-uploaded. [23:56] I'm going to dinner. jacob poke me later for an advocation if you don't get 2 now [23:57] thanks nhandler (/me just had a delicious pizza)