[00:06] <SpamapS> Hmm.. do we need NLS in the wget udeb? that would be another 3k down..
[00:09] <cjwatson> SpamapS: we shouldn't
[00:09] <SpamapS> cjwatson: right, BB doesn't have any either by the looks of it
[00:10] <SpamapS> cjwatson: since I have you here, how do you feel about -Os ? I've never used it.. but it did drop nearly 20kB off the code size (and 10kB off the udeb size) for wget
[00:19] <cjwatson> SpamapS: *shrug*
[00:19] <cjwatson> SpamapS: we use it in a few places in d-i I think
[00:19] <SpamapS> cjwatson: ACK. uploaded a new wget with as many disables and optimizations as I could find in 2 hours.
[00:20] <SpamapS> cjwatson: from 159K to 130K
[00:20] <cjwatson> better than nowt
[00:20] <SpamapS> cjwatson: note that the rebuild just with precise's toolchain raised it from 152K to 159K
[02:46] <slangasek> broder: cool :)
[02:56] <An-iSociaL> whatup! ubuntu is booted with xdesktop on my Motorola Xoom!
[05:32] <pitti> Good morning
[05:32] <TheMuso> Morning pitti.
[05:32] <An-iSociaL> mornin
[06:53]  * SpamapS takes a deep breath and uploads mysql 5.5
[07:50] <broder> bdrung: fyi, it looks like bug #808767 no longer affects the lintian build; i can build 2.5.4 on precise-i386 without error
[08:03] <dholbach> good morning
[09:42] <pitti> hm, today's images grew by 8 MB
[09:42] <sladen> pitti: KDE update last night/
[09:42] <pitti> no, Ubuntu's
[09:42]  * pitti checks
[09:43] <pitti> == Added packages ==
[09:43] <pitti> libicu48 (8.1 MB)
[09:43] <pitti> ah, there we go
[09:43] <pitti> libicu44 still has a few rdepends
[09:43] <pitti> also, it was only 7 MB
[09:44] <pitti> but at least the bulk of it will resolve itself with NBS
[10:19] <micahg> pitti: can I get you do 2 PPA copies for me
[10:19] <pitti> micahg: beer accepted
[10:19] <pitti> micahg: sure, which?
[10:20] <micahg> pitti: ubuntu-mozilla-security, firefox 8.0+build1-0ubuntu0.11.10.3 to ubuntu/primary oneiric (Security) and firefox - 8.0+build1-0ubuntu0.11.04.3 to natty-security
[10:24] <pitti> micahg: done
[10:24] <micahg> pitti: thanks :)
[10:35] <hadess> somebody tell diwic to stop creating surveys for software he doesn't maintain
[11:04] <sladen> pitti: has there been a recent upload to xdg-user-dirs?  bug #893526  has just been filed by a user.  We've seen it before as bug #209513 but what concerns me this time around is wanting to rename ~/./ to ~/Public/  (ie move folders between different levels)
[11:05] <pitti> this question also happens if you change your locale
[11:06] <pitti> sladen: last upload was mid June
[11:12] <cjwatson> pitti: icu> yep, remaining unuploaded piece in main is libreoffice; see last night's conversation ...
[11:13] <pitti> cjwatson: right, seems fine; just wanted to make sure it's not something less transietn
[11:13] <pitti> and yay typing
[12:02] <sladen> pitti: the user claims not to have
[12:04] <pitti> slangasek: is that oneiric or precise? might be a lightdm change
[12:11] <pitti> slangasek, TheMuso: uploaded GTK 3 multiarch-ification, FYI
[12:12] <pitti> TheMuso: so you can probably re-enable it in at-spi again, too
[15:13] <seb128> do we need to file rm bugs for sources that got dropped from debian because they are deprecated and non maintained and buggy?
[15:14] <seb128> or is somebody looking at those and mirroring the action in Ubuntu every now and then? (looking at pessulus in this case)
[15:16] <pitti> occasionally; that's "process-removals"
[15:16] <pitti> seb128: so, no need to file them now
[15:16] <pitti> in fact, I'll take a run through this
[15:16] <seb128> pitti, ok, thanks
[15:20] <pitti> Riddell: libkipi was removed from Debian (debian bug 615579), with "taken over by kdegraphics from kde4"; but we still have rdepends
[15:20] <pitti> Riddell: should we follow suit, or keep the sourrce?
[15:24] <Riddell> pitti: let me investigate
[15:24] <Riddell> upstream aren't entirely consistent there
[15:28] <Riddell> pitti: libkipi has indeed been merged into kdegraphics but now kdegraphics is released as separate split tars so we have libkipi again, so please keep it
[15:28] <pitti> Riddell: ah, ok; so eventually Debian will do the same, I figure
[15:29] <pitti> Riddell: thanks for checking!
[15:29] <Riddell> right
[15:30] <pitti> cjwatson, mterry: I think I fixed the libgail-3-common induced uninstallability now
[15:31] <pitti> on powerpc/armel buildds need to catch up
[15:36] <mterry> pitti, sweet!
[15:37] <pitti> mterry: just in case you look at precise_probs and want to LART me :)
[15:37] <mterry> :)
[15:37] <pitti> in hindsight I should have updated seeds and -meta first, then gtk
[15:59] <Andy80> hi all
[16:01] <Andy80> where can I find informations/tutorials about splitting something in multiple packages? (for example: foo, foo-docs, foo-libs, foo-data ecc....). I've given a look at the official packaging documentation but it looks like this thing is not covered (at least on the doc I was reading it wasn't). Thanks!
[16:31] <SpamapS> Andy80: I'm not sure there are specific guides for that.. but usually you will need to be more explicit with each of the .install/.docs files to specify the package name.. and add sections to the control file for each
[16:32]  * SpamapS changes gears
[16:32] <SpamapS> so, I uploaded a merge of mysql-5.5 from Debian experimental and since it is new to Ubuntu, its in the NEW queue. Does it have to pass a full NEW review or am I just waiting for an AA to ack it?
[16:36] <pitti> SpamapS: as long as it's a sync, we usually just wave it through
[16:36] <pitti> SpamapS: so if it's a merge, an AA will at least have to do some cursory checks
[16:52] <SpamapS> pitti: ACK.. so is there anyone I can poke for an expedite? I need to get the libmysqlclient transition rolling. :-P
[16:52] <pitti> SpamapS: https://wiki.ubuntu.com/ArchiveAdministration#Archive_days says Riddell is the master of the day, but not sure whether that's still current
[16:53] <Riddell> pitti: actually I'm just doing my first archive duties for many months
[16:53] <Riddell> SpamapS: what needs doing?
[16:54] <pitti> Riddell: ah, still know the knobs? :-)
[16:54] <SpamapS> Riddell: mysql-5.5 .. merge from Debian experimental is in the NEW queue.
[16:54] <SpamapS> Riddell: I would be most overjoyed if you could give it a high priority. Thanks! :)
[16:54] <Riddell> SpamapS: then I'll get to it shortly as I do New processing (unless there's a paticular hurry)
[16:55] <SpamapS> Riddell: next few hours is fine. It will take 2 hours to build, and then I have to start a rather large transition for the new libmysqlclient
[16:56] <Riddell> that'll be fun :)
[16:56] <pitti> SpamapS: ugh, ABI change? happy buildd warming :)
[16:57] <Riddell> SpamapS: you'll probably need to poke me again once it's built then to let the New binaries in
[16:57] <pitti> Riddell: JFYI, I'm currently running process-removals (man, looong backlog), so if you try to run some tools and you don't get a lock, that'd be me
[16:57] <pitti> please poke me if you need it
[16:57] <Riddell> ack
[17:02] <SpamapS> Riddell: will do. :)
[17:02] <SpamapS> pitti: what was fun was convincing MySQL upstream that there actually was an ABI change.. they didn't believe it until we showed them the crashes *of their own client software*
[17:03] <pitti> superm1: ah, I can't push to ~mythbuntu/mythtv/mythtv-fixes -- can you please change ttf-droid to fonts-droid and ttf-liberation to fonts-liberation? (presumably some more fonts, too)
[17:03] <pitti> superm1: they recently got renamed
[17:06] <pitti> cjwatson: argh, sorry -- I removed ttf-farsiweb, saw too late that it is used by d-i; it was renamed to fonts-*
[17:09] <superm1> pitti: can you give me a merge request, i'll sort out the permissions problem later and merge it?
[17:09] <superm1> (assuming you already made changes locally of course)
[17:09] <pitti> superm1: not yet, I just saw that I can't push the branch
[17:09] <pitti> but can do
[17:10] <slangasek> pitti: multiarchification - whee, thanks :)
[17:14] <pitti> superm1: https://code.launchpad.net/~pitti/mythtv/font-rename/+merge/83047
[17:18] <pitti> cjwatson: I pushed the change to bzr, but as I haven't done a d-i upload yet I rather don't try it now without your guidance
[18:10] <bryceh> pitti, have you heard of anyone having troubles building pbuilder images for precise?  I posted my log at #893700.
[18:11] <seb128> bryceh, he left for the day but that doesn't seem really pitti specific
[18:11] <seb128> that maybe something the stable team can help with?
[18:11] <seb128> i.e cjwatson or mterry
[18:11] <bryceh> seb128, ok
[18:12] <mterry> y
[18:12] <mterry> o
[18:12]  * mterry looks at bug
[18:13] <bryceh> heya mterry, I poked around to see if anyone got that same error as I but found nothing; wondering if it's some odd perl conflict
[18:13] <bryceh> or something particular to the pbuilder version I have installed
[18:14] <mterry> bryceh, hm maybe cjwatson would know right away, he did the perl transition
[18:14] <broder> bryceh: debootstrap was having issues with perl
[18:14] <broder> perl-modules in particular
[18:14] <broder> cjwatson fixed it with the most recent debootstrap upload to unstable, if you want to grab that
[18:14] <bryceh> broder, ah that sounds like a plausible suspect
[18:15] <broder> (as i recall, both the old and new versions of perl-modules were still getting published, and debootstrap couldn't handle it)
[18:15] <bryceh> broder, ah
[18:16] <bryceh> hmm, not sure if I can direct pbuilder to pull bits from debian, but sounds like we'll be syncing that in before too long?
[18:16] <broder> bryceh: you use your own debootstrap, not the chroots
[18:17] <broder> so you can just install it on your host machine
[18:17] <broder> err, not host, but ykwim
[18:34] <bryceh> broder, awesome, it built the precise pbuilder image now, thanks!
[18:34] <broder> np :)
[18:37] <SpamapS> Crap.. mysql 5.5 FTBFS on armel
[18:37] <SpamapS> https://launchpadlibrarian.net/85702840/buildlog_ubuntu-precise-armel.mysql-5.5_5.5.17-4ubuntu1_FAILEDTOBUILD.txt.gz
[18:38] <SpamapS> /build/buildd/mysql-5.5-5.5.17/sql-common/client_plugin.c:247:5: error: incompatible type for argument 5 of 'add_plugin'
[18:38] <SpamapS> Guessing this is a common failure
[18:52] <Sarvatt> SpamapS: http://bugs.mysql.com/bug.php?id=62769
[18:53] <SpamapS> Sarvatt: nice.. thanks!
[18:57] <jono> hey all
[18:57] <jono> is Precise reasonably stable at the moment?
[18:57] <jono> I am thinking of upgrading
[19:04]  * SpamapS dist-upgraded a week ago but has yet to reboot
[19:18] <bryceh> jono, I've just upgraded one system and fresh-installed 3 others
[19:19] <bryceh> jono, aside from a few assorted glitches (that I think you'd be unlikely to see) it went smoothly
[19:25] <jono> thanks bryceh
[19:51] <Thetawaves> WW: Explicitly asked to ignore failures (probably not good)
[19:51] <Thetawaves> II: New modules (you've been busy, wipe the poop off your nose)
[20:22] <cjwatson> pitti: fonts-farsiweb> that's fine, thanks - no urgency on an upload, it can go into the next one that's needed for some other reason.  I've committed the corresponding seed change too
[20:23] <cjwatson> bryceh: oh, yeah, I meant to sync debootstrap from unstable today for that but forgot (something to do with being on holiday maybe ;-) ).  Synced now
[20:27] <bryceh> cjwatson, thanks.  Should the fix be sru'd, for folks that are doing pbuilder of precise on stable series?
[20:27] <bryceh> cjwatson, if you can point me at the appropriate patch I can do the sru grunt work on it.
[20:28] <cjwatson> bryceh: I commented on the bug.  I don't think it will get through the SRU process before the archive is changed such that it goes away on its own
[20:28] <bryceh> cjwatson, ah ok
[20:28] <cjwatson> bryceh: http://anonscm.debian.org/gitweb/?p=d-i/debootstrap.git;a=commitdiff;h=0fbf86aa8fbcd06cb62fddddcfd4605fef2ee8b2 if you want, though
[20:29] <cjwatson> I could put it in -backports, that would be quicker
[20:29] <cjwatson> (after this sync publishes)
[20:29] <bryceh> cjwatson, ah that's not a bad idea
[20:33] <bryceh> cjwatson, I'll wontfix the natty/oneiric tasks unless you think there's any chance that this could recur such as when q-series opens?
[20:36] <cjwatson> at the very least it will recur at each perl transition
[20:41] <broder> cjwatson: :-/ i recognize that we can move packages faster through -backports, but i'm not super thrilled about using backports for bugfixes
[20:43] <jtaylor> has someone looked at backport sphinx 1.07?
[20:44] <broder> jtaylor: bug #?
[20:44] <jtaylor> none yet
[20:44] <broder> then probably not :-P
[20:44] <jtaylor> I'm asking first, not that I do duplicate work
[20:45] <Laney> everything ought to be on bugs
[20:45] <jtaylor> would be quite useful for some python pacakges using dh_sphinxdoc
[20:45] <broder> searching the ubp project, i see bug #256017 and bug #291825, both of which should be killed off at this point anyway
[20:46] <broder> but nothing recent
[20:46] <jtaylor> k thx
[20:47] <broder> jtaylor: i'd encourage you to play with requestbackport in ubuntu-dev-tools, either from precise or from ppa:udt-developers/daily
[20:48] <tumbleweed> broder: oh, I never added existing backport searching to requestbackport
[20:48] <broder> tumbleweed: yeah, but i'm slowly ramping up to encouraging people to always using it for opening new bugs anyway
[20:48] <cjwatson> broder: Well, it's not so much that as that historically I've kept debootstrap up to date in -backports anyway because lots of people want the new scripts symlink
[20:49] <broder> cjwatson: i thought we sru'd those...
[20:49] <cjwatson> Somebody started SRUing them without asking me first
[20:49] <cjwatson> I'd already been backporting them for some time before that
[20:49] <Laney> i only see it in updates for lucid
[20:50] <broder> heh, ok. i have no problem if there's precedent
[20:50] <cjwatson> So basically it's a mess
[20:50] <cjwatson> Alternatively somebody could upload the relevant bits of KDE to make perlkde buildable and then the problem will go away for now :-)
[20:51] <cjwatson> In principle perl transitions should be completeable in about two or three days; less if we had good enough tools to stage them somewhere
[20:51] <Laney> It does seem more appropriate to cherry-pick that fix for SRU, but I recognise that it may be more work than it's worth. So I'm happy with carrying on keeping debootstrap more-or-less up to date in backports..
[20:52] <Riddell> I am in the proccess of merging and uploading large parts of KDE so it should happen soon
[20:53] <cjwatson> it's just smokekde and okular I think, if I can convince you to prioritise :-)
[20:53] <Riddell> I'll see what I can do
[20:54] <cjwatson> cool
[21:01] <TheMuso> pitti: Thanks./
[21:09] <jtaylor> cjwatson: you merged svn last, are you considering fixing bug 881862 in oneiric?
[21:09] <jtaylor> it would be shame when the good triaging of the reporter goes to waste
[21:10] <jtaylor> the issue should be fixed in precise with the merge
[21:42] <RainCT> pitti: Hey. Any idea why the Traceback of bug #760111 & dupes show "/home/username" (ie. with "username", literally) as the path?
[21:57] <seb128> RainCT, some users consider their login name to be private info, apport replace it for privacy reasons
[21:57] <RainCT> seb128: it's still in the bug titles, though
[21:57] <seb128> "bug" ;-)
[21:57] <seb128> the intend is to hide it
[21:58] <seb128> the title case was not considered I guess
[21:58] <RainCT> seb128: Okay, thanks for the info. So it's not XDG_HOME being messed up or something :P
[21:58] <seb128> no it's not ;-)
[21:59] <seb128> "    def anonymize(self):
[21:59] <seb128>         '''Remove user identifying strings from the report.
[21:59] <seb128>         This particularly removes the user name, host name, and IPs
[21:59] <seb128>         from attributes which contain data read from the environment, and
[21:59] <seb128>         removes the ProcCwd attribute completely.
[21:59] <seb128>         '''
[21:59] <seb128> "
[21:59] <seb128> RainCT, ^ that's what apport is doing basically (that's from report.py in the apport source code)
[22:30] <bdrung> broder: k, synced
[23:16] <szymon_g> hi
[23:18] <szymon_g> hm... why does ubuntu have 2 installation-cd: "desktop" and alternate? wouldn't it be easier to keep it on one (since space isn't that big problem- it will be to big for cd anyway...)?
[23:26] <penguin42> szymon_g: Don't forget the desktop is also a live cd
[23:27] <RAOF> szymon_g: You don't need both the alternate and desktop installers; they're two different ways to arrive at the same installed system.
[23:29] <szymon_g> penguin42, yes, i know. i wonder- why they are seperated: on the start of system /when you choose like "check cd for defects" "check memory" etc etc/ someone could choose: either livecd /and maybe simple install from it/ or "advanced" way- i.e. textual installer, with lvm/raid etc
[23:30] <szymon_g> so it would be easier for some people /like me: no need to get 2 images: livecd for checking how unity has been changed + alternate for installing on lvm/raid/
[23:30] <TheMuso> szymon_g: That used to be available in the DVD version.
[23:30] <szymon_g> "used to be" :?
[23:30] <szymon_g> was dvd ceased?
[23:30] <TheMuso> As of oneiric, I believe the DVD version in that form is no longer available.
[23:30] <RAOF> They're separated because you can't fit both of them on the same CD.
[23:31] <TheMuso> There is still a DVD version, but its a live system only, and much smaller.
[23:31] <szymon_g> RAOF, but the next ubuntu wont be on CD either
[23:31] <RAOF> It will be if it fits :)
[23:31] <szymon_g> and textual installer + lvm/raid support wont occupy much space
[23:32] <TheMuso> Yes it will actually, you need all the debs of the system to be installed.
[23:32] <szymon_g> RAOF, i heard rumors that it will be approx 750mb o.O
[23:32] <RAOF> Yes, it will.  it will occupy 700 MB of space. :)
[23:32] <RAOF> There is essentially nothing shared between the livecd and the alternate installer.
[23:33] <szymon_g> so- are there any chances to introduce lvm/raid configuration in livecd installer?
[23:33] <RAOF> Of course.
[23:33] <RAOF> It's a long-standing feature request.
[23:33] <TheMuso> But not a priority afaik.
[23:33] <RAOF> I don't think that anyone's currently working on it, but if you wanted to I don't know of any reason why good patches wouldn't be accepted.
[23:34] <szymon_g> RAOF, i'm just asking. my only programs were "hello world" in C. the most beautiful 32 lines of code you could ever see
[23:34] <szymon_g> ;)
[23:35]  * Frostbite pats szymon_g
[23:35] <szymon_g> :) btw, is anyone working on delta-deb?
[23:35] <szymon_g> it's huge bandwith saving /my record is 96% of savings/