=== soren_ is now known as soren === almaisan-away is now known as al-maisan [05:34] * micahg wonders how tumbleweed got cairo-dock-foo to build === al-maisan is now known as almaisan-away [08:06] This is proberly a silly question but if I wanted to package a Java program could I just folllow the Packaging Guide? [08:07] sure [08:08] geser: Thanks [08:08] just make sure that you recompile any .jar files instead of using the prebuild upstream ones [08:09] Ok, thank you very much [08:50] micahg: given its dependencies, it gave me no trouble (they're in NEW) [08:50] tumbleweed: ah, I looked :) [08:50] didn't see them there, but I trust you [08:51] otoh, bug 923688 is pretty high on the annoyance list atm [08:51] Launchpad bug 923688 in bzr-builddeb (Ubuntu) "bzr crashed with AttributeError in tree_unapply_patches(): 'DirStateRevisionTree' object has no attribute 'last_revision'" [High,Triaged] https://launchpad.net/bugs/923688 === directhe` is now known as directhex === almaisan-away is now known as al-maisan [11:31] so how many hours till feature freeze, or is it frozen already? [11:31] 2100UTC [11:31] thx [11:32] requestsyncs from Debian NEW are not possible, right? [11:32] I don't think they are, because it's not accessible [11:34] well, the only way they certainly are [11:34] s/only/old/ [11:35] if you've got the source package itself, then it's possible [11:36] you can dget from NEW [11:36] oh you can now? [11:36] * ajmitch guesses it's been this way for awhile :) [11:37] * micahg thinks you can at least [11:52] ubuntu new, yeah [11:53] actually that might be the case in which you still need dgetlp [11:53] there is still one, I forget what it is [11:53] Laney: no, was referring to Debian NEW [11:53] but debian new, no I don't think so [11:54] even if I ssh to ries I can't access the files [11:59] hmm, must have been confusing it with something else [12:10] you can get thnigs from incoming.debian.org, but that's not NEW [12:10] and of course, many packages are in VCS... === Tonio__ is now known as Tonio_ [13:11] I would much prefer people to wait and FFE than do fakesyncs [13:12] * tumbleweed too [13:13] btw, any opinions on the tesseract FFe? Its in the grey area for me [13:13] much better OCR, but have to remove a reverse-dep [13:14] will look later [13:20] * Laney begins writing an annoucement of the uninstallable packages challenge [13:20] http://debian.titanpad.com/7 [13:22] err, requires login [13:22] i made it public [13:22] and I am editing it without being logged in right now [13:23] aha [13:23] yah boo [13:23] nigelb deleted his blog post for the ftbfs jam [13:23] oh, no, just broke old links [13:24] (or the link in tumbleweed's original email was wrong :P) [13:25] * tumbleweed refuses to beleive that possibility [13:25] hmm [13:25] the graphs aren't that convincing that the jam made a difference [13:27] we were at ~300 FTBFS packages at the time... [13:28] as in [13:28] there's a steady downward trend that seemed to start before the jam happened [13:29] we got one or two new people in, working on it. And they mostly evaporated again [13:30] * tumbleweed discovers a qa-ftbfs cronjob still running on his server, and kills it [13:32] is there a better link for oneiric-historical.html? [13:32] no, I couldn't find one [13:33] didn't know if it moved to ubuntuwire [13:33] it did, but not at that time [13:33] http://people.ubuntuwire.org/~stefanor/lp-ftbfs-report/historical/ [13:33] huh, sub-50? [13:34] primary archive, not rebuilds [13:34] i suppose the archive rebuild would be worse [13:34] but still that's better than i would have necessarily thought [13:34] wgrant isn't using my graph additions in his rebuild-ftbfs-test reports [13:35] (poke poke) [13:37] I don't think we should do automated bug filing for the uninstallability issues, just for the jam. So collaborate with an etherpad? [13:38] ye [13:46] Hi all, having a few issues updating a package for Precise, build log is here https://launchpadlibrarian.net/93036528/buildlog_ubuntu-precise-i386.ruby-rvm_1.10.2-0ubuntu4_FAILEDTOBUILD.txt.gz, i am getting permission errors but not sure where to look to get to the bottom of the problem [13:48] you can't write to the 'real' filesystem when building a package; make the build system install to pkgbuilddir/debian/tmp [13:48] adhorden: it's no trespecting the DESTDIR provided [13:49] also, it should not be installing to usr/local at all [13:49] it's probably a broken install script in this case [13:54] Laney: this is what I was thinking the install script seems to be breaking things here, might have to hack that up so that it uses DESTDIR [13:55] yes, that is the standard interface [13:55] I was thinking this would be an easy script to package up, its been nothing but headaches [13:56] life is made much easier when people use standard build systems [13:56] ruby packages also seem to have that affect on people [13:57] i have still managed to avoid it completely [13:57] * tumbleweed too [13:57] yup, and in this case rvm does not use a standard build system it uses its own install shell script [13:58] I thought iamfuzz promisdet o look after rvm? (I seem to remember being hesitant to want it in the archive at all) [14:01] not sure who the maintainer is, there was a bug in the package so I took a look into it and bumped the upstream version as it was really out of date, I was going to submit my patches to bring it upto date [14:02] adhorden: ubuntu packages don't have maintainers, and this one needs some maintainance [14:02] ah right, well I am taking a look into it as I am using precise on my systems at present and would be nice to have a native working package for rvm === al-maisan is now known as almaisan-away === almaisan-away is now known as al-maisan === al-maisan is now known as almaisan-away === arand_ is now known as arand [15:47] tumbleweed: re: bug 408757, I talked to the upstream maintainer and they don't have a problem with what you suggested. I already submitted the patch into Debian, working on updating the debdiff [15:47] Launchpad bug 408757 in nsca (Ubuntu) "init sscript has no Usage option" [Undecided,Confirmed] https://launchpad.net/bugs/408757 [15:48] it's actually bug 897006 but the debdiff is in the other issue (trying to close both of them in one shot( [15:48] Launchpad bug 897006 in nsca (Ubuntu) "Build nsca-client package" [Undecided,Confirmed] https://launchpad.net/bugs/897006 [16:01] we've got through "grave" in rcbugs. It'd be nice if we could get through all the serious bugs that are fixed in new upstream versions, but that's a lot for one evening :P [16:29] Laney: I did? Looking [16:29] nigelb: nah, I think it was a wrong link in the initial mail [16:29] ah, quite possible. [16:30] Laney: http://nigelb.me/ubuntu/2011/06/27/fix-ftbfs-jam.html [16:31] the link said /07/ [16:32] draft maybe? [16:33] Hrm, interesting. [16:51] ok, I'm through all the low-hanging serious-severity rcbugs that are fixed with new upstream versions (well, one still test-buliding) [16:52] poor non-x86 buildds [18:05] tumbleweed: nice work with the rc bugs [18:07] tumbleweed: so you did decide to sync yaws in the end? I wasn't sure what the changelog was meaning by removing erlang patches [18:14] so, time for some last minute merging ;) [18:15] still a couple of hours let I think [18:15] s/let/left/ [18:16] yes 21:00 utc I was told [18:17] * ajmitch is still getting his laptop back to usable [18:18] meh scipy built a couple minutes to early and did not pick the new numpy :( [18:18] on armel [18:22] ouch [18:22] how long did it take to build? [18:23] 4 hours [18:23] scipy numpy less [18:25] rebuilds of -ubuntuX just increase X? or add a buildY to the version? [18:28] merge, merge! [18:29] hm ok how do I merge a 13.5 mb package with my 11kb/s upload connection? [18:29] I could sync it and then add the diff in a second upload ;) [18:30] do you have a time machine at hand? :) [18:30] hm can I dput from alioth? (after I signed it locally?) [18:31] what's the latest possible time i could file a sync request and get it merged into 12.04 without exception? also 21:00 utc or does it need to be processed already by that time? [18:32] jtaylor: I don't see why you shouldn't be able to dput from alioth [18:33] stefanct: my understanding is that we will be accepting sync requests without FFes if they are filed before FF [18:35] am i allowed to file a sync request even though the debian package has not hit the archives yet? (yes i am quite desperate :/) [18:36] if its new ffe's are usually easier to get [18:36] ic [18:36] we are preparing a release upstream atm... but i doubt that it gets ready in time [18:37] oh its only an update not completely new? [18:37] what are you updating? [18:37] flashrom, app to flash bios chips etc. last release was last summer [18:37] jtaylor: not a completely new package, no [18:42] Hey. i is possible to patch one package? [18:49] don't want to rock the boat, but another other sponsors care to comment on bug 408757 [18:49] Launchpad bug 408757 in nsca (Ubuntu) "init sscript has no Usage option" [Undecided,Confirmed] https://launchpad.net/bugs/408757 [18:49] and bug 897006 [18:49] Launchpad bug 897006 in nsca (Ubuntu) "Build nsca-client package" [Undecided,Confirmed] https://launchpad.net/bugs/897006 [18:50] pabelanger: do you have any sense of when the debian maintainer is planning to release their new version with the split? [18:52] broder: not sure, just pinged him but waiting to hear back. [18:53] The package is not very active, last release was Dec. 2010 [18:53] Now that I am using it more for a customer site, I suspect I'll be pushing upstream to help keep it active [18:55] pabelanger: oh, i see. the client just isn't getting shipped at all currently? i assumed that you were taking a monolithic package and splitting it up into smaller packages [18:56] i'm much less opposed to what you're doing [18:56] (splitting up a package - especially when you know debian's about to do it too - has enough risk of introducing subtle differences between debian's split and ubuntu's split that it always makes me uncomfortable) [18:57] pabelanger: anyway, the patch seems conceptually reasonable, but i can't do a full review right now [18:57] Okay, thanks for the look [19:55] hm I want to sync the new plplot, but it requries a mini transition [19:55] 3 packages [19:55] one of them ftbs but fix available in debian testing [19:56] should I file a sync request instead of syncing that? because plplot will not be done building until freeze [20:00] jtaylor: Is the fix for testing a new upstream version? If not, you are fine, sync plplot now and do the rebuilds after plplot finished on all architectures. [20:00] it is [20:00] new upstream [20:00] though a minor, not sure how extensive [20:07] It isn't too hard to get a FFe shortly after freeze, especially when it's only a minor update. === chrisccoulson_ is now known as chrisccoulson [20:48] phew all done [20:49] all remaining numpy packages had new versions in debian..., so I pulled them all hopefully I don't have to much mess to clean up this weekend xD [20:50] (they all have very extensive test suites which were successful, so its unlikely) [20:57] ajmitch: I probably should have tested it, but it built... [20:57] jtaylor: Re should you wait on syncing stuff: There's a long (unfortunate) tradition of tons of crap getting uploaded right at feature freeze, so I wouldn't suggest handicapping yoursef. [20:58] don't know if tulp is worth sneaking in. There's a much newer upstream verison in debian which builds unmodified, but maybe not on all archs... [21:00] broder: the nsca client is currently being shipped, just in the server pakage [21:00] *package [21:00] oh, i see [21:00] well, it's still using pre-existing packaging to do the split === medberry is now known as med_ [21:03] the pre-existing rules allows building a client-only package [21:03] but if it's going into the archive, I say let's split it properly... [21:04] so, feature freeze in effect now? [21:04] yup, but FFes are cheap for the first few days :) [21:07] ah, I see someone hit the Debian NEW queue this evening [21:14] tumbleweed: so we can bug you or any other friendly release team member? :) [21:14] that's what we're here for (/me hides) === Laney changed the topic of #ubuntu-motu to: REVU is back up | Precise: Feature Freeze | Want to get involved with the MOTU? https://wiki.ubuntu.com/MOTU/Contributing | Sponsor queue: http://bit.ly/fz6AyQ | http://qa.ubuntuwire.com/ftbfs | http://people.canonical.com/~ubuntu-archive/NBS | http://qa.ubuntuwire.com/bugs/rcbugs/ [21:17] Laney: thanks :) [21:17] DENIED [21:17] :'( [21:17] * ajmitch shall go back to lurking instead [21:17] ♥ [21:18] * Laney → pub [21:18] * tumbleweed just got back from the pub [21:18] bit early for the pub for me [22:41] hi, hey how can I pass the -j option (dpkg-buildpackage) via debuild... [22:41] I did it.. but can remember how... [22:41] -j8 directly just fails. [22:41] works for me [22:42] jtaylor, debuild -us -uc -j8 [22:42] something similar? [22:42] yes [22:42] it passes it to dpkg-buildpackage which has the option in its manpage [22:43] dpkg-buildpackage -rfakeroot -D -us -uc -j8 failed [22:43] if I remove the -j8... I see it begins the compilation [22:43] which package? [22:43] no further information? [22:44] shouldn't be related to any package [22:44] general question. using this with our packages. [22:44] will use dpkg-buildpackages directly [22:45] oh well.... even the dpkg-buildpackages fails. will build slowly then [22:52] hm how do I fix this: http://paste.ubuntu.com/845099/ [22:54] a figured it out, had to remove a -build-deps