=== medberry is now known as med_out [07:48] good morning === chrisccoulson_ is now known as chr1sccoulson [10:35] ARGH [10:35] http://people.canonical.com/~ubuntu-archive/transitions/ghc.html [10:51] I should have predicted this :( :( :( [10:51] * Laney mashes the keyboard in irritation [10:51] For a second there, I read tradition. [10:52] when you maintain GHC it does become rather traditional, indeed [11:02] Laney: do ghc transitions ever end? [11:02] this one was about to [11:02] and then I uploaded GHC again last night [11:02] I have the impression that at least one package on any arch has always a wrong dependency and needs a rebuild [11:02] imagine yourself travelling around a circle :-0 [11:02] :-)* [11:03] * DktrKranz patpats Laney [11:03] * Laney shelters underneath DktrKranz [11:03] you have the power of 'wb nmu' [11:04] nope [11:04] 'you' as in 'debian' [11:04] makes such frustrations go away somewhat :P [11:04] definitely [11:05] wasn't "rebuild package" button in progress? [11:05] not that i've heard [11:07] DktrKranz: not "rebuild package" but a "sync package" button is worked on [11:08] https://dogfood.launchpad.net/ubuntu/oneiric/+localpackagediffs [11:11] cool [11:12] yes, really interesting === med_out is now known as med === med is now known as medberry === BackFromTheUnder is now known as Slick === Slick is now known as Ekx [17:33] Vortexirc looks for linux/chat interested people, if you want to join irc.vortexirc.com 6667 see ya :D === Ekx is now known as Ekx-Off [17:35] don't do that [17:59] broder: regarding seamonkey, we'll need to push 2.1 when it's released to the stable releases, so idk if an SRU is worthwhile [18:00] 2.1 uses the newer hunspell [18:00] oh, really? yeah, in that case it doesn't seem like it makes much sense to keep working [18:01] yeah, the 3.5 branch which seamonkey 2.0 is based on will be EOL on June 21, so there will be no more releases for it [18:02] * micahg probably should've commented on that bug [18:56] bdrung: I never got an answer about syncing eclipse from experimental, I did a test build on amd64 on oneiric and it was fine, I would do one more with i386 before requesting the sync since i386 failed before with gcc-4.6 [19:04] What went wrong there? Shouldn’t it be possible to change ghc’s debian/control without breaking all the ABI hashes? [20:18] Hi. I'm stuck with updating a package. I've already uploaded the .orig.tar.gz (2.3) version to ppa, and I later discovered there was a missing dependency, so I edited debian/control, but I can't reupload the .diff.gz file because the same version of the diff file (2.3) was uploaded and launchpad won't accept same version but different content. [20:19] and if I pump the number in debian/changelog, I can't seem to get an updated .diff.gz (to 2.3ppa1 for example), I have to rename the orig file 2.3ppa1 as well which is wrong [20:19] so how do I fix that, what am I doing wrong? [20:20] upload without the orig source [20:20] KNRO: it sounds like you've done a handful of things wrong [20:21] KNRO: you should use a debian revision (i.e. 2.3-0ppa1 [20:21] debuild -sa [20:21] KNRO: first, if this isn't a "Debian/Ubuntu-native package" - i.e. the software meaningfully exists outside of Debian/Ubuntu, you should add a Debian revision number, like micahg said [20:22] it's not native to Debian/Ubuntu [20:22] but also, while you're experimenting in PPAs, you should add something to the version number - something like ~ppa1 - that leaves you room to change the version number in your PPA [20:22] KNRO: right, so the version of the Debian package shouldn't be "2.3", it should be "2.3-0ubuntu1" [20:22] then you add on the PPA tag (so "2.3-0ubuntu1~ppa1") [20:23] okay, and the .orig filename is then what? [20:23] 2.3.orig.tar.gz ? or 2.3-0ubuntu1.tar.gz ?? [20:30] the furst [20:30] s/u/i/ [20:32] so if the orig file was foo_2.3.orig.tar.gz and my debian version is 2.3-0ubuntu1~ppa1, I can't run debuild -sd or -sa or any combination without renaming the .orig file!!! [20:32] But the .orig file name should remain the same!! [20:32] KNRO: what does debian/source/format say if anything? [20:33] I see ... dpkg-source: warning: no source format specified in debian/source/format, see dpkg-source(1) [20:33] guess that's my problem... [20:34] KNRO: that's the debian version set in debian/changelog? [20:34] in debian/changelog I have 2.3-0ubuntu1~ppa1 [20:35] if it can't find the orig then you'll get a native package [20:35] there is no source/format under debian [20:35] don't worry about that [20:35] how do I tell it to look for foo_2.3.orig.tar.gz then? [20:35] and take the diff from that? [20:36] show us the complete output from 'debuild -S' please :-) [20:37] ok [20:38] here http://pastebin.com/cE3xGsyj [20:40] ls ../libapogee2_2.3.orig.tar.gz [20:40] KNRO: dpkg-source is using the version number 2.3ubuntu1~ppa1, not 2.3-0ubuntu1~ppa1 [20:40] the hyphen is critical [20:41] Laney: yes it exists... ? [20:41] yeah, you need to fix that version :-) [20:42] ohhhhh [20:42] 2.3 then hypthen [20:42] okay let me try [20:46] it worked!! Thanks guys, this was driving me insane === yofel_ is now known as yofel [21:04] another question, if I have a package already built in maverick, and I want to build it for natty ppa, is it possible to build without updating the changelog file? [21:07] no [21:07] you can't have different binaries from the same revision [21:08] even for different distributions? [21:08] you can binary copy it from maverick to natty in the PPA without changing the versions but thats usually a bad idea and can't rebuild it that way [21:09] actually it's often not a bad idea [21:09] that's what happens for new ubuntu series after all [21:10] I'll just update the changelog file and do it the easy way [21:11] try copying the binaries and see if it works === roaksoax__ is now known as andreserl [23:16] micahg: i answer yesterday or so. the answer was: yes, please sync it. [23:16] bdrung: ah, sorry, must have missed it, xchat doesn't always tell me when I have a message waiting, thanks, will do over the weekend === randalogger is now known as info [23:17] micahg: np. i would have done it yesterday if requestsync wouldn't crash === info is now known as infologger