[05:46] <dholbach> good morning
[05:46] <fabrice_sp> morning
[05:48] <dholbach> hiya fabrice_sp
[05:48] <fabrice_sp> Hi dholbach
[05:49] <fabrice_sp> isn't it a bit early for you also?
[05:49] <fabrice_sp> :-)
[05:49] <dholbach> 6:50 here
[05:49] <fabrice_sp> The same here
[05:50] <dholbach> I get so excited in the morning because of work with all you guys that I can't go back to sleep ;-)
[05:52] <fabrice_sp> \o/ dholbach
[06:08] <porthose> dholback: ping
[06:08] <dholbach> porthose: I'm "dholbach" :-)
[06:08] <dholbach> porthose: what can I do for you?
[06:09] <porthose> during your busy day today
[06:09] <porthose> would you sponsor an FFe for me?
[06:09] <dholbach> porthose: is the FFe ACKed already?
[06:09] <porthose> yes +2
[06:09] <dholbach> great, bug number?
[06:10] <porthose> #263405
[06:10] <dholbach> bug 263405
[06:13] <dholbach> porthose: I saw that in the sponsoring queue for a while and thought "ah yeah, that's emgent's apache-related project - why doesn't he take care of it"..... seems that I'm mistaken (rapache <-> ampache)
[06:13] <dholbach> I'll take a look at it in a bit
[06:13] <porthose> k thanks :)
[06:36] <dholbach> porthose: uploaded
[06:36] <porthose> dholback: thanks :)
[06:36] <dholbach> "dholbacH" :-)
[06:37] <porthose> aaahhh ooppps sorry :-[
[06:37] <dholbach> don't worry - I just wondered, why you don't use tab-completion there :-)
[06:38] <porthose> didn't know about it, will have to read more  about it:-[
[06:39] <dholbach> which IRC client do you use?
[06:39] <dholbach> typing   "dhol<tab>"    should auto-complete the name for you
[06:39] <porthose> oohh ok
[06:40] <porthose> pidwin I think the one that comes with hardy
[06:40] <dholbach> ah ok
[06:42] <porthose> hey it's 0100hr here give me a break the brain is dead :)
[06:42] <dholbach> I wasn't complaining :)
[06:42] <porthose> Im still brain dead hehe :)
[06:43] <porthose> dholbach:  thanks and good night :)
[06:46] <wgrant> dholbach: Do people mispronounce your name like that IRL too?
[06:46] <dholbach> wgrant: "Holbach" is not very easy for native English speakers
[06:46] <wgrant> I hope I have it right.
[06:46] <dholbach> "holback" is pretty common there
[06:46] <wgrant> Ew.
[06:47] <dholbach> I don't mind :)
[06:47] <dholbach> I guess they have more problems with "Sébastien Bacher" :)
[06:47] <dholbach> everybody says "Sebastian" and doesn't bother to pronounce "Bacher" at all :)
[06:48] <dholbach> or just "Seb"
[07:04] <jml> dholbach: isn't there a lot of variation in Germany for pronouncing 'Holbach' though?
[07:17] <dholbach> jml: there's just two groups of people, some say "Hohlbach" (long o) and others say "Hollbach" (very short o)
[07:18] <persia> dholbach: Are there not also variations in pronunciation of 'ch': while it's rarely 'k', I'd expect that it would move from the back of the throat through the middle of the mouth, depending on region.
[07:18] <jml> what about the 'ch'? I've been told that Kölners make it sound like the English "sh"
[07:32] <dholbach> persia: "Bach" is a pretty common word - it's "brook" in English, so everybody gets it right
[07:33] <persia> dholbach: Oh, I was sure it was "right", just for different values of "right".  An example from English would be the wide variety of ways to pronounce "creek"
[07:38] <dholbach> jml: "ch" can be "ç" or in the Köln region a bit more like "ʃ", but also "ʜ" (the throat sound in the "ch" of "Bach")
[07:38] <dholbach> I hope I got the phonetics right
[07:38] <dholbach> this is how I pronounce my name: http://www.youtube.com/watch?v=JJzM2LNOtWU :-)
[07:47] <jml> dholbach: is this a cunning trick to get me to upgrade some packages?
[07:48] <dholbach> jml: :-)
[07:49] <jml> http://www.youtube.com/watch?v=VyEl3w7SFK4&feature=related is actually more my speed :)
[07:49] <dholbach> jml: I should have taken the "MOTU - Getting Started" one.... you're right :)
[07:49] <dholbach> hehe
[07:50] <jml> I now have at least three motives for becoming a MOTU, I should really get my act in gear.
[08:50] <iulian> Good morning.
[09:57] <pochu> hi all
[09:57] <pochu> is there anybody from motu-release?
[10:02] <TheMuso> pochu: Yep.
[10:02] <pochu> hi TheMuso
[10:03] <pochu> TheMuso: is it neccessary to write a bug report for every upload saying it's bug fix only, attaching the changelog entry? I thought it wasn't neccessary but it's been pointed to me that https://wiki.ubuntu.com/FreezeExceptionProcess says that (under the motu-release bug fix only section)
[10:04] <TheMuso> pochu: I can't remember off hand, I jsut refer to the docs in cases where I am not immediately sure.
[10:05] <pochu> TheMuso: ok, I'll mail ubuntu-motu@ to clarify it
[10:05] <Hobbsee> pochu: if you don't, i doubt you'll get caught up regardless.
[10:05] <pochu> TheMuso: thanks
[10:05] <persia> pochu: I've always felt that that clause only applied in cases where there was a new upstream version, to confirm, but it's leftovers from when we had an explicit UpstreamVersionFreeze
[10:07] <pochu> I don't think writing a bug report has any value if I can upload without anybody reviewing it
[10:07] <pochu> it's just an annoyance
[10:08] <Hobbsee> pochu: i agree with you, but got overruled.
[10:10] <norsetto> pochu: if you write it in the changelog entry is enough, bug report is only necessary for things that do not get uploaded directly (like syncs, or sponsoring requests)
[10:11] <persia> pochu: Filing a bug report for something you're fixing has value for documentation purposes.  It allows one to provide significantly greater verbosity about the issue than one would want to include in debian/changelog
[10:12] <persia> Note that it's not appopriate in all cases, but can be useful in some.
[10:15] <pochu> norsetto: I see. in those cases (e.g. sync request) is it necessary to put the upstream changelog diff, or just saying it's a bug fix only release is good enough?
[10:15] <norsetto> pochu: the more you document the more is obviously better
[10:16] <persia> pochu: The basic rule of thumb is to provide enough documentation that nobody can complain that you did it wrong later.
[10:16] <iulian> RainCT: ping
[10:17] <pochu> persia: that makes sense
[10:17] <norsetto> pochu: what persia said can also be said in shorter words "CYA", a useful acronym in real-life (tm) too
[10:18] <pochu> heh
[10:18] <persia> Well, it's also interesting for changelog readers.  I read changelogs, and I much prefer to see quick snippets, and only read the details if they interest me.
[10:19] <persia> I don't like the habit of some people pasting all of NEWS into the changelog: it no longer makes narrative sense.
[10:19] <pochu> I still would like to be able to request syncs or upload new revisions or whatever without having to put a changelog.diff in a bug report, so could we update the policy? :)
[10:20] <persia> pochu: It's considered best practice to put a debian/changelog diff in a sync request, even when we're not in a freeze.
[10:20] <persia> For new upstream revisions, it's not hard to copy the content into the description when filing a bug, and you'll want to cover yourself anyway.
[10:21] <pochu> persia: right, I'm talking about the upstream ChangeLog, sorry
[10:21] <persia> For most uploads, the debian/changelog is going to get posted to several relevant bugs anyway.
[10:21] <norsetto> pochu: and don't get yelled at by those nasty archive admins
[10:21] <pochu> heh, in this case it was a contributor :)
[10:21] <persia> pochu: I think that belongs in the bug, so I can view it easily without having to download and extract the package if I have questions.
[10:22] <pochu> but he was backed up by policy, so I can't complain ;)
[10:22] <persia> pochu: Also, especially because of the hack in CDBS to not ship upstream changelogs, I need to get it somewhere.
[10:22] <pochu> right
[10:22] <persia> pochu: So, while I understand you don't want to post it, please do, because I might want to read it.
[10:23] <persia> (and that's as a *user*, not as a developer)
[10:23] <pochu> so if consensus is to have it there, I'll try to do it
[10:23] <persia> I'm not sure it should be in policy: just because I want to read it doesn't mean you shouldn't have a policy debate, but it's at least a use case for having it.
[11:00] <iulian> git show 5fe08835
[11:00] <iulian> Ooups
[11:00] <iulian> Fail
[11:05] <pochu> mail sent about FFe
[11:16] <didrocks> hi, I try to find from which package /usr/include/netinet/in.h is installed from
[11:16] <didrocks> $ dpkg-query -l /usr/include/netinet/in.h give me "no package corresponding to ..."
[11:16] <didrocks> I don't really see why :)
[11:17] <torkel> didrocks: dpkg -S /usr/include/netinet/in.h
[11:18] <didrocks> what's the difference beetween the two commands?
[11:19] <torkel> didrocks: dpkg-query -l wants a package name not a file in the package
[11:19] <torkel> s/the/a
[11:19] <didrocks> yes, I misread the man, sorry :/
[11:19] <didrocks> thanks a lot torkel, it worked :)
[11:21] <norsetto> didrocks: you may want to use apt-file if you don't have the package installed
[11:22] <didrocks> norsetto: apt-file search <file>, isn't it?
[11:22] <norsetto> didrocks: indeed
[11:22] <didrocks> norsetto: thanks, very useful :)
[11:28] <persia> Note that apt-file may be out of date (as Contents.gz isn't updated with everypublisher run), and often doesn't work early in a cycle (as there is yet no Contents.gz).
[12:44] <nhandler> Is anyone else here receiving an error message when they try and install kdebase-runtime?
[12:57] <iulian> Wow - octaviz takes ages to build!
[12:58] <iulian> On my lappy of course, no idea if it takes so much time on other machines.
[15:23] <nxvl> bash-completion is the package in charge of making all the non-commands completions? or there is another package?
[15:23] <Hobbsee> nxvl: command-not-found
[15:24] <persia> nxvl: You'll want to look in all of bash, bash-completions, and command-not-found
[15:24] <nxvl> what i want is to type 'apt-get inst' then tab and it completes 'install'
[15:24] <nxvl> is that what those packages are for?
[15:25] <nxvl> oh!
[15:25] <persia> nxvl: I think that's in bash-completions, but some completions are in bash directly.  Look at the packages to determine which is correct for apt-get
[15:25] <nxvl> command-not-found is that awesome magical package that tell me what to install to get that command?
[15:25] <nxvl> awesome!
[15:25]  * nxvl HUGS Hobbsee and persia 
[15:26]  * persia doesn't like command-not-found because it keeps making suggestions for "les" when I'm not sure whether I want ls or less.
[15:32] <superm1> persia, make an alias in your bashrc for les -> less
[15:32] <superm1> persia, that's what i did a log time ago for sl
[15:33] <slytherin> nxvl: bash-completion is the one you are looking for. And many packages install completion logic. ex pbuilder.
[15:39] <_ruben> superm1: got a question about dkms .. im failing to find out if dkms can be used for distributing binary packages .. all 'docs' ive seen on it, show "build-on-install" .. and i dont want to have to build it on all machines i want a certain 3rd party kmod
[15:39] <superm1> _ruben, yes you can distribute prebuilt kernel modules with it
[15:39] <soren> Which package do I file FFe's against when it's about a new package?
[15:39] <superm1> _ruben, there are options when building to only include the binary or only the source instead of both
[15:39] <soren> Just file it against "ubuntu"?
[15:40] <_ruben> superm1: hmm .. must've been missing that then
[15:40] <_ruben> superm1: which doc/link/whatever do you suggest for getting started with dkms?
[15:40] <superm1> _ruben, but be wary of shipping with just the binary module, particularly because newer kernel ABI's are added to Ubuntu.
[15:41] <superm1> _ruben, a combination between the manpage and https://help.ubuntu.com/community/Kernel/DkmsDriverPackage
[15:41] <tillux> heya there. is there a way to create a gtk/whatever gui in order to replace an existing ncurse one?
[15:41] <_ruben> superm1: ah, hadnt checked the man page (shame on me) .. the h.u.c url looks familiar though
[15:42] <superm1> _ruben, and http://linux.dell.com/git/?p=dkms.git;a=blob;f=debian/HOWTO.Debian;h=532e5b9d996e97d678eddfe807a79a746abcd6ba;hb=HEAD gives a small overview
[15:42] <superm1> that howto is in the package's /usr/share/doc directory of course
[15:42] <superm1> but the man page will tell you a lot of options that aren't discussed in other howtos
[15:43] <superm1> so yeah shame on you :)
[15:43] <ssaboum> hi everyone
[15:43] <_ruben> superm1: i'll go hide in shame now :)
[15:48] <ssaboum> hi rain
[15:49] <ssaboum> i worked on the branch of revu about email subscription
[15:49] <ssaboum> if you want to check it out
[16:01] <norsetto> RainCT ^^
[16:06] <bddebian> Heya gang
[16:11] <compengi> hello all. i'm interested in learning how to build dep packages and joining your team. i followed a video guide from here https://wiki.ubuntu.com/MOTU/Videos everything went just fine, untill i was going to run "debuild -S -sa" i received an error debuild: fatal error at line 1247: dpkg-source -b ed-0.9 failed. any ideas?
[16:18] <directhex> need the complete error log
[16:18] <directhex> in isolation that doesn't say enough
[16:21] <compengi> okay. 1 second
[16:22] <compengi> directhex, http://paste.ubuntu.com/46274/
[16:26] <norsetto> compengi: looks like you have a (editor backup ?) binary file left into debian
[16:28] <compengi> norsetto, i don't really know what you mean. i was just following the steps as the video showed
[16:28] <nxvl> has someone see this error before -> http://paste.ubuntu.com/46277/
[16:29] <slytherin> compengi: check if there is any file with extension .swp in debian/ directory
[16:29] <compengi> norsetto, and in debian there is only those files "changelog  compat  control  copyright  rules"
[16:30] <norsetto> compengi: how did you edit changelog? The log you showed us show that your have a binary file called debian/.changelog.swp which should't be there
[16:30] <norsetto> compengi: thats becuase its a dot file, use ls -a to see it
[16:30] <compengi> norsetto, yeah true. it's there. why was it created?
[16:32] <norsetto> compengi: how would I know? My guess is that its a backup file but its just a guess
[16:32] <compengi> norsetto, i see.. and should it be deleted?
[16:32] <norsetto> compengi: yes
[17:46] <ykphuah> ﻿after I do apt-get source foopackage, ﻿how do I let apply the debian/patches and do a configure, but stop there, without doing make
[17:52] <slytherin> ykphuah: depends on what patch system is used. If it is dpatch then inside source do 'dpatch apply all' I am not sure about other patch systems.
[17:52] <slytherin> ykphuah: You can find out which patch system is being used using command 'what-patch'
[17:53] <ykphuah> slytherin: it uses quilt, so I have generated a patch of my own... which should be applied after all the patches in debian/
[17:53] <ykphuah> slytherin: and now I want to test out the patch.
[17:54] <ykphuah> slytherin: so I just quilt push -a, apply my own patch, then do dpkg-buildpackage?
[17:55] <asomething> https://wiki.ubuntu.com/PackagingGuide/PatchSystems#quilt
[17:55] <slytherin> ykphuah: I don't understand why are you trying to apply patch manually. But yes, quilt push -a should do.
[17:55] <asomething> ykphuah: with quilt there's a couple extra step
[17:57] <ykphuah> ok, actually, I fixed one of the bugs/annoyances in network-manager, what I did from beginning is, do "apt-get source network-manager", then "dpkg-buildpackage", this will make sure my source directory is applied with those debian/patches, and configured with the correct flags, and then I start editing the source codes.
[17:57] <ykphuah> after I edit the source codes, I do make and make install for testing.
[17:58] <ykphuah> and after I finish doing testing this way, i generated a patch.
[17:58] <ykphuah> move the path I modified as "new", then I do "apt-get source network-manager" again, move that to "old", then I do quilt push on the "old"
[17:58] <ykphuah> and do a diff, take out the parts that I changed.
[17:58] <ykphuah> now I end up with a simple .patch file, which I would like to test.
[17:59] <ykphuah> I have went thru the URL for PatchSystems before, but I haven't really find a page that tells me how to simplify what I described above.
[17:59] <ykphuah> Was I doing it totally wrong way?
[17:59] <ykphuah> i wanted to submit this patch to the launchpad for review.
[17:59] <slytherin> ykphuah: I am not entirely sure that procedure is correct. I suppose dpkg-buildpackage will deapply all patches once it is done with build.
[18:02] <ykphuah> so is there a way to ask dpkg-buildpackage stop halfway?
[18:02] <ykphuah> i assume it does a couple of steps -> patch, configure, make, make install
[18:02] <ykphuah> is there a way to get it to stop halfway? i have read the man pages but couldn't seem to know how.
[18:03] <slytherin> ykphuah: What do you exactly want to do? Test your patch right?
[18:04] <slytherin> ykphuah: I have a doubt that the way you created your patch is right.
[18:04] <ykphuah> slytherin: i want to learn what's the proper way of creating a patch for reviewing in launchpad.
[18:04] <slytherin> ykphuah: But I am no expert in quilt so please wait for someone else to answer.
[18:04]  * slytherin likes dpatch more.
[18:05] <laga_> quilt is teh awesome IMHO. dpatch is a bit clumsy ;)
[18:05] <ykphuah> slytherin: well, so dpkg-buildpackage knows to call quilt or dpatch?
[18:05] <persia> !patch
[18:05] <laga_> although i should use its patch editing more
[18:06] <persia> ykphuah: There are a number of different ways to produce patches in the right format.  The link ubottu just provided is probably the best overview.
[18:06] <slytherin> laga_: quilt is too complex for me.
[18:06] <persia> In the specific case of network-manager, it's probably best to present your fixes as a bzr branch.
[18:06] <persia> (this is not typically true, but that package is special)
[18:06] <laga_> slytherin: i used it for a big patch series of about 150 patches
[18:07] <laga_> and it was awesome for that :) no packaging task, tho
[18:07]  * ykphuah goes off to read the two links
[18:07] <slytherin> ykphuah: Or probably talk directly with asac on #ubuntu-devel to discuss the fix.
[18:09] <sebner> huhu persia :)
[18:10] <slytherin> What is the procedure for request upload permission for a specific package?
[18:11] <persia> slytherin: Apply to the technical board.  It's never yet been approved except in cases where MOTU have access to specific packages in main, or for a single kernel developer.
[18:11] <slytherin> persia: what is the mailing list of technical board?
[18:12] <persia> slytherin: technical-board@lists.ubuntu.com
[18:12] <slytherin> thanks
[18:31] <ScottK-laptop> brandon|work: Do you still have the new upstream tarballs of endeavour2 and avscan?  Upstream has disappeared and I'll need those to be able to backport clamav from Intrepid to Hardy.
[18:44] <brandon|work> ScottK-laptop, possibly, my dev computer crashed a couple weeks ago, but I might have them in backups
[18:44] <brandon|work> if not, I can remake them
[18:51] <ScottK> brandon|work: That'd be great.  Now that 0.94 is in the Intrepid archive (and in Main) and the libclamav4 -> libclamav5 transition is done, I want to start working on the backport.
[18:52] <brandon|work> ok, I will get it up by the end of the weekend (sooner if I have the backups)
[18:55] <ScottK-laptop> Thanks.
[20:24] <karooga> hi, how do I get someone to review on revu?  does it happen automatically?  or do I need to be patient for someone to take a look after FeatureFreeze?
[20:26] <geser> karooga: you need to be very patient
[20:26] <laga_> yeah, unfortunately :/
[20:28] <karooga> geser: no prob.  Docs online just don't specify if it's of the polling or interrupt variety. :-)
[20:29] <geser> karooga: do you have an exception to get this package into intrepid? if no, then don't expect much review before the intrepid release
[20:32] <karooga> geser: no no exception for intrepid.  (Just a first time package / packager...)
[20:47] <norsetto> geser: how important is it that we do this switch to openjdk-6 for intrepid ?
[20:54] <geser> norsetto: I don't know if it's a release goal (do we have any for universe?). But openjdk-6 is the default JRE so it would be really nice if the -java package use it (especially when they use the non-free sun-java-* packages now)
[20:55] <norsetto> geser: sure, I understand its nice, I'm just trying to understand the benefit of allowing this versus the potential of breaking things
[20:56] <norsetto> geser: is there a need to patch upstream sources for this?
[21:00] <geser> depends on upstram but afaik no, either it compiles with it or not (if it doesn't compile yet it still build-depends on sun-java-*)
[21:03] <geser> norsetto: but I'm no java expert, I more a upload bitch for slytherin :) He knows more about java packaging than me.
[21:04] <norsetto> geser: ok, then please talk with him because I think it would be sensible to only sync with Debian when no new upstream version is involved, and to patch our packages when it is (or, justify properly the FFe, which I don't find is the case for those he raised so far).
[21:08] <geser> norsetto: will do
[21:10] <norsetto_> geser: sorry, got disconnected, have I lost anything?
[21:14] <geser> norsetto_: no, I just wrote that I will talk with slytherin
[21:14] <norsetto_> geser: ok thx, I really appreciate your help
[21:17] <Adri2000> norsetto_: hi
[21:17] <norsetto_> Adri2000: hi there
[21:19] <Adri2000> norsetto_: regarding your ngircd sru verification, the update fails only if ngircd is not running right?
[21:20] <norsetto_> Adri2000: indeed, after a reboot for instance
[21:21] <norsetto_> Adri2000: I must say, I haven't tested if it would work when the daemon is running, but as long as /var/run/ngircd is there it should work
[21:25] <Adri2000> norsetto: yeah, it works when it is running. that's how I tested
[21:25] <Adri2000> norsetto: anyway thanks for catching, I'll fix that now
[21:26] <norsetto> Adri2000: thx to you; btw my patch is just a proposal, you may obviously find something better
[23:09] <nhandler> Is anyone else here receiving an error message when they try and install kdebase-runtime?
[23:09] <goshawk> hi
[23:09] <ScottK> nhandler: #kubuntu-kde4 is probably a better channel for that question.
[23:10] <goshawk> if someone has free time, can please review this? http://revu.ubuntuwire.com/details.py?package=dsss
[23:10] <nhandler> Thanks ScottK