[00:07] <virhilo> slangasek: nope, in naty 64 i have to export them any time when compiling somehting
[00:07] <slangasek> compiling how?
[00:08] <virhilo> by ./configure and make:)
[00:08] <slangasek> yes, but using what compiler?
[00:08] <slangasek> this all works fine with the default gcc compiler in natty
[00:09] <virhilo> slangasek: gcc
[00:09] <slangasek> well, I can't reproduce the problem you're describing
[00:09] <virhilo> slangasek: it was not able to fiund libpng zlib and few others
[00:10] <virhilo> slangasek: try compiling python 2.7 for example:)
[00:10] <slangasek> ah - that has nothing to do with gcc
[00:10] <virhilo> or pil for python
[00:10] <virhilo> slangasek: it uses gcc
[00:10] <slangasek> no, that's an issue with python 2.7's library detection
[00:10] <doko> virhilo: which python version?
[00:10] <virhilo> 2.7
[00:10] <virhilo> slangasek: ahhh...
[00:11] <slangasek> the upstream library detection code inspects the filesystem directly; a 'gcc -lpng' or 'gcc -ljpeg' works just fine
[00:11] <virhilo> slangasek: so i should report bugs for them?:)
[00:12] <virhilo> doko: Python 2.7.1 (r271:86832 to be exact:)
[00:12] <slangasek> this has been discussed upstream already, and mentioned in barry's post to ubuntu-devel here: https://lists.ubuntu.com/archives/ubuntu-devel/2011-April/033049.html
[00:12] <doko> virhilo: try to build from the current branch
[00:14] <virhilo> doko: you mean directly form hg from default branch?
[00:15] <doko> virhilo: from the 2.7 branch, or rebuild the oneiric package on natty
[00:17]  * virhilo trying
[00:21] <virhilo> clonning slow:P
[00:27] <virhilo> doko: works:) awesome
[01:47] <lfaraone> Would having update-manager source its mirror lists from the web be a useful feature to implement?
[01:48] <lfaraone> I noticed that in Lucid, several mirrors don't even exist anymore. (washdc-linux.com was squatted)
[01:48] <ohsix> i thought it did already
[01:49] <ohsix> the mirror chooser tries to connect to all of them just the same; to see if they work and how fast/close they are, so bad ones won't hurt much
[01:49] <lfaraone> ohsix: /usr/share/update-manager/mirrors.cfg is where the mirror list is sourced from.
[01:50] <ohsix> ok
[01:53] <lfaraone> ohsix: the mirror picker does do a download test, but people manually picking will choose bad mirrors sometimes.
[02:33] <persia> lfaraone, I believe there was a discussion at UDS about an entirely new and better way to select mirrors.  I may be mistaken, but I'd encourage you to review the specification list just in case (unless someone gives you a good answer sooner)
[03:28] <micahg> @pilot in
[05:20] <pitti> Good morning
[05:21] <ajmitch> morning pitti
[05:50] <micahg> hi pitti , I'm looking at bug 775900 for SRU, I think the idea is to upload the version in oneiric to natty, I was thinking of just adding a changelog entry on top with the version 2.0.0-0ubuntu0.1, WDYT?
[05:52] <pitti> micahg: if it's actually a backport, why not 2.0.0-1~lucid1?
[05:52] <pitti> but I don't mind much either way
[05:53] <micahg> pitti: ah, yeah, it's for an SRU
[05:53] <pitti> right, but technically it's still a backport
[05:54] <pitti> no need to set X-Ubuntu-Maintainer etc.
[05:54] <micahg> pitti: hmm, I just compared a debdiff of the archive versions and it doesn't exactly match the debdiff in the bug...
[05:58] <pitti> micahg: diff that causes problems? I. e. new dependency, major changes, etc? or just autoconf noise?
[05:58] <micahg> pitti: idk, supposedly he diff'd the same versions I did
[06:01] <micahg> I'm going to use the version in oneiric since it's more authoritative one, since the bug suggests that was the goal
[06:03] <pitti> micahg: sounds fine
[06:07] <pitti> hm, what's up with the publishing -- evolution 3.0 was built 10 hours ago, and http://archive.ubuntu.com/ubuntu/ still has 2.32
[06:08] <pitti> it's not in NEW
[06:12] <lifeless> pitti: see #is
[06:15] <pitti> lifeless: I'm there now
[06:15] <pitti> ah, known issues
[06:15] <pitti> thanks
[06:17] <micahg> pitti: is it possible to use the auto backporter or should I just upload with -sd -v and the extra changelog?
[06:17] <pitti> micahg: I think upload it manually, that's easier; just append a new changelog and use -v -sa
[06:18] <pitti> erm, -v -sd, sorry
[06:18] <micahg> pitti: k
[07:04] <dholbach> good morning
[07:11] <micahg> @pilot out
[07:14]  * dholbach hugs micahg
[07:16]  * micahg hugs dholbach 
[07:46] <pitti> doko: good morning
[08:21] <cjwatson> lfaraone: https://launchpad.net/ubuntu/+spec/foundations-o-apt-mirror-method
[08:36] <doko>  o debian-archive-keyring: debian-archive-keyring
[08:36] <doko>    [Reverse-Recommends: debootstrap]
[08:36] <doko> cjwatson: ^^^ you're an uploader, what to do?
[08:37] <doko> Sweetshark, pitti: could you have a look at the hyphen* component mismatches?
[08:41] <doko> pitti: same for ttf-sil-nuosusil
[08:51] <RAOF> cjwatson: I'd appreciate a review of http://tinyurl.com/3sm2rlv with respect to bug #742683 if you've got time.
[09:04] <lool> Hey
[09:05] <lool> I'm puzzled that seabios 0.6.2-0ubuntu1 was built in oneiric and accepted, isn't in NEW, yet doesn't show up in rmadison's output or in the archive
[09:05] <lool> Launchpad thinks it was published 10 hours ago
[09:07] <lool> ah I see in backlog that we have some publishing issues
[09:14] <pitti> lool: #is says ETA in about an ohur
[09:14] <pitti> hour, too
[09:16] <lool> yup
[09:18] <lool> hmmm 133 packages to satisfy tomboy bdeps, ouch
[09:21] <cjwatson> doko: ignore that entry in c-m, I'll do something about it in debootstrap
[09:25] <Sweetshark> doko: will do
[09:30] <hrw> pitti: thx for fix
[09:32] <cjwatson> RAOF: yes, I think that's fine.  You could simplify by using sed -i
[09:32] <cjwatson> well, fine as in terrifying
[09:47] <RAOF> cjwatson: Heh.  I'm following the idiom of console-setup there.  Yes, moderately terrifying.
[09:59] <cjwatson> doko: adjusted in Debian to be sensitive to the host distribution; next autosync will take care of it
[10:00] <doko> thanks!
[10:01] <doko> cjwatson: any idea about fast-tracking bug #785607 and bug #785609 ?
[10:06] <cjwatson> doko: well, I'm not on the MIR team, but lib*-perl packages rarely seem problematic
[10:07] <hrw> pitti: pkg-create-dbgsym works great for g++ now
[10:07] <pitti> nice!
[10:07]  * hrw -> breakfast
[10:17] <ev> pitti: I wasn't able to attend the networking session at UDS. Is network-manager 0.9 (instead of, say, conman) the way forward?
[10:17] <ev> the specification seems to indicate so, but I wanted to confirm: https://blueprints.launchpad.net/ubuntu/+spec/desktop-o-desktop-network-enhancements
[10:18] <ev> as I'll be using the dbus api in ubiquity to construct a wireless connection page
[10:18] <pitti> ev: that's my impression anyway, also after talking with cyphermox
[10:18] <ev> great
[10:18] <ev> that makes me a happy bunny
[10:18] <pitti> ev: I don't know of a plan to switch to connman by default in oneiric
[10:18] <ev> good deal
[10:18] <pitti> . o O { after all, this stuff works.. }
[10:19] <ev> indeed
[10:19] <ev> this wasn't something I heard recently, I just recall conversations a release or two ago about eventually moving to conman
[10:20] <ev> and wanted to be sure that wasn't still on the cards
[10:21] <pitti> might have been a temporary fashion
[10:21] <pitti> but I don't know much about connman TBH
[10:21] <pitti> if NM moved to system-wide connections by default, it'd do everything I want from it :)
[10:22] <ev> I suspect it was exactly that
[10:22] <pitti> it does suppor them, it just defaults to by-user
[10:31] <cjwatson> freeflying: can libofetion be synced from Debian?  It's in the OpenSSL 1.0.0 transition list
[10:33] <freeflying> cjwatson: that would be great
[10:33] <cjwatson> freeflying: have you checked that there are no outstanding Ubuntu changes?  if so, file a sync request?
[10:33] <cjwatson> ScottK: could you merge boost1.46 from Debian?  It would fix the libtorrent-rasterbar build failure
[10:34] <freeflying> cjwatson: will do it
[10:35] <cjwatson> thanks!
[11:48] <kenws> Hi there, I'd like to rebuild the firefox deb with an additional compiler flag. From looking at the files in the debian directoy it seems that debian/mozconfig would be the appropriate one, correct? Another option might be to just set the C[XX]FLAGS in debian/rules. Any suggestions are appreciated.
[12:05] <kenws> hm, setting DEB_CFLAGS_APPEND might be another way...
[12:38] <ScottK> cjwatson: I'll try to have a look at it a bit later today.
[12:38] <cjwatson> thanks
[14:47] <ScottK> SSD is handy when working on boost.
[14:49] <slangasek> also, opium
[14:50] <ScottK> That too.
[14:59] <hallyn_afk> pitti: I've sent a new source package to the libcap2 debian maintainer in the hopes that he'll push it and we can sync (to fix an open bug in both debian and ubuntu).  Is there any problem with waiting and syncing, as opposed to pushing it to oneiric myself?
[15:01] <ramvi> Is there an apt configuration which tells Ubuntu to choose package-ubuntu packages instead of debian packages?
[15:05] <ScottK> ramvi: The only way to do this is not have Debian repositories in your sources.list.  This is also highly recommended as Ubuntu makes no attempt at all to maintain binary compatibility with Debian.
[15:09] <broder> ev, pitti: I asked cyphermox about connman at one point during UDS. He said that it was still a pretty good ways from having feature parity with NM
[15:11] <cjwatson> doko: http://orangesquash.org.uk/~laney/transitions/ocaml.html - could you update llvm-2.7 and llvm-2.8?  I tried to do a trivial rebuild of llvm-2.7 but it failed
[15:11] <cyphermox> broder: ev: those who want to use connman can already do so by installing connman/indicator-network and it works okay for simple requirements (wifi, wired, etc)
[15:12] <broder> cyphermox: Right, but we won't be switching to it by default anytime soon
[15:12] <doko> cjwatson: how urgent is this? would rather try to move to llvm-2.9
[15:15] <cjwatson> doko: I suppose that depends on how urgent it is that libllvm-ocaml-{2.7,2.8}-dev are uninstallable
[15:15] <cjwatson> doko: I'd like to complete at least one of the four transitions I'm looking at this month, though ;-)
[15:15] <doko> hmm, it's a leaf package
[15:16] <doko> really need to work on some blueprint stuff
[15:16] <ev> broder: yeah, I'm quite happy with n-m.
[15:16] <cjwatson> it's certainly not today-urgent, I just don't really want it to be my problem as I'm not familiar with those packages
[15:17] <cjwatson> it might just be a matter of adding <stddef.h>
[15:22] <ogra_> can someone from the sru team reject https://launchpad.net/bugs/721147 ? i need to re-upload (accidentially catched the wrong patch)
[15:22] <ogra_> err
[15:22] <ogra_> https://launchpad.net/ubuntu/natty/+source/flash-kernel/2.28ubuntu20 i meant
[15:36] <hyperair> kirkland: pin
[15:36] <hyperair> g
[15:37] <apw> doko, you got any idea how much longer the arm builds are likely to take of gcc ?  wondering when to come back and check :)
[15:47] <hrw> apw: give it 8-10h
[15:50] <apw> hrw, 8-10 more!  blimey
[15:50] <apw> its been going some 24 already
[15:54] <ScottK> cjwatson: boost1.46 uploaded.
[15:54]  * ScottK will upload boost-mpi-source1.46 here in a bit.
[15:57] <cjwatson> thanks
[16:01] <bryce> any archive admin feel like nuking nvidia-graphics-driver-180 from the archive?  (LP: #756197)
[16:03] <slangasek> I'm all about nuking packages
[16:03] <bryce> :-)
[16:04] <GunnarHj> cjwatson: Hi Colin, any chance that you can review/sponsor a very small im-switch merge + upload to natty-proposed? It's a regression thing, and the proposed change is important to input method users, so the natty-proposed part is kind of urgent. https://code.launchpad.net/~gunnarhj/ubuntu/oneiric/im-switch/lp-663776/+merge/61683
[16:06] <cjwatson> GunnarHj: sure, looking
[16:08] <cjwatson> GunnarHj: perhaps change xinput.d/default as well?  (maybe only in oneiric, since it's a conffile so might prompt merges)
[16:11] <GunnarHj> cjwatson: The reason I didn't include it is that that is just a description of how it used to be, and not part of the actual code.
[16:13] <cjwatson> you don't think it's effectively documentation?
[16:13] <cjwatson> hmm, I suppose not really
[16:14] <GunnarHj> cjwatson: I'm not sure. Happy to change default also if you really want me to.
[16:15] <cjwatson> GunnarHj: I honestly think this is a bit bizarre though
[16:15] <cjwatson> LC_CTYPE = character type => obvious choice for input method handling
[16:15] <cjwatson> LC_MESSAGES = translation selection => nothing to do with input methods
[16:15] <cjwatson> are you sure the real bug isn't in language selector for failing to set the right set of locale elements?
[16:18] <GunnarHj> cjwatson: I thought of that; and maybe it is (and/or GDM). But changing the set of env. variable in l-s and gdm is a bigger thing. Could we possibly consider this a hack that is good enough for now, and reconsider l-s + gdm later on (and maybe reverse the hack)?
[16:20] <cjwatson> GunnarHj: sure, as long as that's borne in mind and recorded somewhere
[16:20] <GunnarHj> cjwatson: Promise I'll file a bug about it. :)
[16:21] <cjwatson> thanks
[16:22] <cjwatson> GunnarHj: uploading to oneiric - I'll do a natty-proposed upload in your name too, if that's OK?
[16:22] <GunnarHj> cjwatson: Of course it's ok. Thanks!
[16:27] <cjwatson> GunnarHj: done and done
[16:28] <GunnarHj> cjwatson: Great, thanks a bunch! Going to write that bug report...
[16:52] <mterry> doko, perl MIRs done.  phew!
[16:53] <mterry> doko, I do love doing perl MIRs though.  All the packaging is the same and up to date, all have tests, all pretty simple
[18:02] <doko> mterry: thanks! promoted
[18:58] <doko> apw: it's now in the archive (because I forgot to re-enable the testsuite)
[19:45] <hallyn> jdstrand: so the libvirt folks say phy was never supposed to work for qemu, it's only valid for xen domains.
[19:46] <hallyn> Daviey: ^  jdstrand was saying this might mess up eucalyptus, that they use 'driver name='phys'' for their xml
[19:47] <hallyn> so, shall i add a patch to allow phy for one cycle, or will that just make upstreams be lazy and not fix it?
[20:02] <cjwatson> james_w: does it really make sense for the importer to continue to run on obsolete series (or at any rate to continue to raise MPs for divergences)?
[20:02] <james_w> cjwatson, is this related to rxtx?
[20:03] <jdstrand> hallyn: interesting
[20:03] <cjwatson> james_w: I've deleted the mails now, but I've been seeing a bunch of MPs for edgy/feisty/gutsy; yes, I think rxtx sounds familiar
[20:03] <SpamapS> pitti: pign, whenever you have a moment.. I have a few people waiting on the squid SRU for lucid.. thanks. :)
[20:04] <SpamapS> hah..pign.. latin for "mispelled ping"
[20:04] <james_w> cjwatson, right. I think that the collisions could still happen on old series. They will only be raised if there are new uploads. There are bugs that are causing false positives now though, I think we should fix those and get a robust and useful service
[20:09] <andyrock> hi all :)
[20:10] <andyrock> is Sebastien Bacher here?
[20:12] <chrisccoulson> andyrock, no, i think he's finished for the week now
[20:13] <andyrock> his nickname is seb128?
[20:13] <jdstrand> cjwatson: hi! is there any policy for adjust $HOME/lintian on cocoplum?
[20:13] <jdstrand> s/adjust/adjusting/
[20:14] <andyrock> chrisccoulson, his nickname is seb128?
[20:14] <chrisccoulson> andyrock, yes
[20:16] <andyrock> chrisccoulson, thanks! is here someone else who is in charge of gtk3 packaging?
[20:16] <chrisccoulson> andyrock, well, the desktop team :)
[20:17] <chrisccoulson> you'd probably be better off asking whatever question you have in #ubuntu-desktop
[20:17] <andyrock> ok thanks!
[21:03] <Daviey> hallyn, Argh!  I think that is a question that really needs upstream comment.
[21:20] <hallyn> davidcalle: which upstream?
[21:20] <hallyn> davidcalle: sorry, not you :)
[21:20] <hallyn> Daviey: which upstream?
[21:21] <Daviey> hallyn, A question for euca, to see how they would react losing the phy0 interface
[21:22] <Daviey> If it's no longer supported by libvirt, i'm loathed to carry support for it if there are other options.
[21:22] <hallyn> they claim it was never supported :)
[21:22] <hallyn> but documentation online made it seem that way imo, so just pulling the rug out wasn't nice
[21:22] <Daviey> okay, if upstream break an unsupported interface :)
[21:22] <Daviey> (upstream libvirt in that context)
[21:22] <ohsix> whats involved in getting something in the contrib repo? (like java and skype is)
[21:22] <hallyn> of course ideally i'd first make sure they do use it,
[21:23] <hallyn> but i can't look at the source code iiuc
[21:23] <maco> ohsix: psst "partners" repo
[21:23] <maco> actually, partner, singular, now i think about it
[21:23] <ohsix> yea that
[21:23] <ohsix> vmware breaks all the time and it'd be cool if it was in there
[21:24] <maco> ohsix: http://www.canonical.com/about-canonical/partnerships
[21:24] <maco> given the name, i think that'd be the page
[21:24] <ohsix> okie dokie
[21:25] <maco> Parallels used to be in there
[21:25] <maco> oh, still are a partner...
[21:25] <maco> hmm, VMWare is listed as a partner already
[21:26] <maco> but that just may be about them supporting VMWare running on ubuntu
[21:26] <broder> ohsix: vmware breaks because they're not keeping up with kernel changes fast enough
[21:26] <ohsix> ya i know
[21:26] <broder> and that's not really connected to whether or not vmware is in the partner repo
[21:26] <maco> yeah, thats pretty much it
[21:26] <broder> they'd just want to put it in until they fixed it
[21:26] <maco> 7.1.4 supports 10.10 now that 11.04 is out! oh goody!
[21:26] <broder> i thought 7.1.3 had the requisite patches...
[21:27] <maco> yeah it ran fine but was unsupported
[21:27] <maco> only 10.04 was supported
[21:27] <broder> anyway, i thought 7.1.4 had the patches to build on 11.04
[21:27] <broder> oh, nope - i'm still patching to get it to work
[21:27] <maco> there were patches to make it build with 11.04's kernel since like january
[21:27] <ohsix> hm
[21:27] <maco> it didnt work with 11.04's GTK though
[21:28] <ohsix> maybe a package that'd build the modules that it needed :[
[21:28] <ohsix> there used to be a package that did that but they were way too old and they started removing some of them
[21:28] <broder> maco: wait...really? ...because i have 7.1.4 working on 11.04 using the system gtk
[21:28] <maco> 7.1.4 was released a few weeks before 11.04 came out, finally including patches to work with the new GTK and the kernel, so it'll run on 11.04, but it's not *supported*
[21:28] <broder> ah, ok
[21:29] <maco> up through the end of march, you had to find the right thread on the vmware forums telling you what to do with ldd to make it go
[21:32] <ohsix> hm
[21:32] <ohsix> it's just a drag; it's one or the other when you're testing alphas/betas
[21:33] <ohsix> they should have a place in the installer that can apply patches before it builds the thing, and patches semi-easily available
[21:33] <ohsix> thanks for the info about 7.1.4, going to try it out
[21:36] <ohsix> it'd be cool to have something like googleearth-package for vmware too; maybe it could be patched there
[21:39] <ohsix> hm iy certainly couldn't work the same, vmware has the bundles behind a login gate; blah
[22:28] <cjwatson> jdstrand: it's probably worth attempting to bring it up to date with current lintian at some point, but make sure to keep a backup copy ...
[23:51] <jdstrand> cjwatson: ack-- for now I just added oneiric, but noted