[00:02] Hey rickspencer3, robert_ancell. [00:02] TheMuso, hi [00:03] robert_ancell: TheMuso: hi guyses [00:04] Hey rickspencer3. [00:04] * awe waves [00:04] hi awe [00:04] https://wiki.ubuntu.com/DesktopTeam/Meeting/2009-06-30 [00:05] I haven't had a chance to copy over any notes from this morning [00:06] I don't actually think it will take to long for us to run through these [00:06] ready? [00:06] go [00:06] sure [00:06] +1 [00:06] ok, first, if you don't have travel arranged for Dublin, please arrange it asap, prices are going up quickly [00:07] A bunch of us are going to Desktop Summit [00:07] Its just about arranged on my part, since I'm travelling with StevenK. [00:07] seb128, Riddell, robert_ancell, me, kenvandine [00:07] TheMuso: great [00:08] I'm battling the travel agent to finally send me an eticket [00:08] hmm [00:08] robert_ancell: do you need help? [00:08] rickspencer3, I've got the route chosen, I just send a reminder email today to confirm that it is booked [00:08] ok [00:09] I'll be traveling starting tomorrow, so let randa know if you have problems getting your travel arrangements closed [00:09] sure [00:09] ok [00:10] so then there was the partner update [00:10] essentially U1 file synching is still almost in universe again [00:10] Cool. [00:10] I guess there was a big round of reviews on the packaging, and this sent dobey back to relayout the bzr branches and such [00:11] also, DXE hasn't quite published their plan for Karmic ... [00:11] but they are doing a new FUSA to be compatible with new gdm [00:12] Is that FUSA going upstream? [00:12] so kenvandine is in charge of getting gdm-new and the new Fusa into the archive soon, and smoothley [00:12] Cool [00:12] * TheMuso notes his activity report is not yet on that pagre. [00:12] robert_ancell: well, it's open source, so if they want it, however, I think it's kind of a replacement, so the intention was not to be a new upstream FUSA (so far as I can tell) [00:13] TheMuso: right, I haven't copied those over yet, I'll get to it before I leave, don't worry [00:13] rickspencer3, I ask because I did some work on the fusa package and it is full of Ubuntu patches and doesn't seem to be worked on upstream. So it would be good if this can replace the existing upstream project [00:13] * TheMuso is not worried. [00:14] robert_ancell: ok, please bring that up with seb128 [00:14] will do [00:14] thanks [00:14] ACTION: robert_ancell to discuss upstream of DXE FUSA applet wtih seb128 and kenvandine [00:14] Kubuntu ... [00:15] first, they have a bootable usb drive making tool now, so they can make netbook images [00:15] second, those images aren't actually tweaked for netbooks yet, but the package of netbook settings is created, it just needs the settings [00:16] so your netbook can be Kubutulicious in a matter of a couple of weeks, I'm guessing [00:16] cool [00:16] third, their was some discussion regarding Kubuntu and web browsers [00:17] Arora is the new KDE default browser, but they can't just remove conqi because they are afraid of backlash and such [00:17] I assume that uses webkit. [00:17] correct [00:17] TheMuso: yes, I believe Arora does [00:18] but not Konqueror [00:18] so in terms of x, they are still ahead on merges, and KMS is on by default, but not causing a big spike in bugs [00:18] Yeah I know konq didn't. [00:19] there are some x freezes associated with suspend/resume, though [00:19] * TheMuso should test his hardware to be sure all is ok. [00:20] it looks like translations are moving along ok [00:21] then we had a big discussion about PPAs versus Archives [00:21] essentially, there has been a tendency lately to put new stuff into PPAs instead of into Karmic universe [00:21] this has caused some unpredictability around scheduling ... as it is less clear when something has "landed", and [00:22] right [00:22] I may reconsider the konqueror issue, depends on the reaction from upstream on changing browser, there's a definate appetite for moving to something newer now [00:22] it also means that we are all developing on difference configurations [00:22] hi Riddell! [00:22] Do you guys have anything in PPAs that are bound for universe or main? [00:22] or know of anything? [00:23] nope [00:23] no [00:23] Well pulse may be tested in a PPA, but it may just go straight to karmic. [00:24] TheMuso: please discuss with pitti when he wakes up [00:24] in essence, we are preferring to go strait to Karmic [00:24] rickspencer3: He already asked me about it. [00:24] heh [00:24] and? [00:24] Well it requires yet another dependency in main before it can go in. [00:25] SO short term, it will be PPA, till we get the new dependency into main. [00:25] KDE 4.3 RC 1 is in PPA and being uploaded to main [00:25] I see [00:25] ok [00:25] TheMuso: please focus on moving it into main as fast as reasonably possible [00:25] Riddell: ok [00:26] rickspencer3: I will, gotta get it packaged first. [00:26] TheMuso: understood [00:26] The point is that we should strive to not things languish in PPAs, the sooner they get into archives the better [00:26] Yep. [00:27] so we shouldn't feel a false sense of security if something is in a PPA [00:27] (this is a general point, not directed specifically at Pulse Audio) [00:27] speaking of PPAs - gdm-new [00:27] * TheMuso nods. [00:27] if you can, please test it from the desktop team ppa [00:28] similarly, please set up asac's mozilla security updates ppa so those can be tested as well when they are available [00:29] here's a link to the mozilla ppa: https://edge.launchpad.net/~ubuntu-mozilla-security/+archive/ppa [00:29] do you guys have any agenda items? [00:29] No. [00:30] no [00:30] No, I do want to follow up with TheMuso afterwards re: my review of his spec... [00:30] ok [00:30] thanks guyses! [00:30] thanks [00:32] awe: Ok, sounds good. As I said, I don't understand the volume scaling stuff at all. [00:32] TheMuso: I have a fair large amount of comments for you... should I add email them to you, or add them to the blueprint whiteboard? [00:33] TheMuso: I think from what I heard at UDS, the volume scaling would be solved by using the Pulse volume control and leaving the ALSA volume fixed [00:33] at least that's what I thought I heard Daniel say [00:33] awe: Right, add them to the whiteboard is probably the best, so then everyone else like dtchen then sees them. [00:34] ok [00:34] One thing's that's confusing is whether I comment on the Blueprint, the Specification or both? [00:34] awe: Thanks [00:35] np [00:35] Well I wasn't the one who proposed the blueprint/spec for karmic, so I don't know. [00:35] TO me, audio is one of those things that requires ongoing maintenance, and when things are made to work upstream, we get them to work in Ubuntu. [00:35] i.e doesn't really require a spec. [00:36] but I'll go with what has been asked for. *sighs* [00:37] ok [00:38] I'll just go ahead and add to the whiteboard as you suggested... [00:42] Thanks. === tkamppeter_ is now known as tkamppeter === asac_ is now known as asac [07:18] Good morning [07:37] hey pitti. Do you know the status of GNOME 3 in Ubuntu? [07:40] robert_ancell: seb and a few others are packaging gnome-shell and zeitgeist [07:40] pitti, are there packages available? [07:40] robert_ancell: as for the library deprecation, seb works on some smaller issues (droppling libglade, etc.), there were some 2 uploads for this [07:40] robert_ancell: no idea, I'm afraid; zeitgeist does have pacakges somewhere [07:41] pitti, also, do you remember a gnome-scan discussion at UDS? What was the outcome/ [07:42] robert_ancell: https://blueprints.edge.launchpad.net/ubuntu/+spec/desktop-karmic-gnomescan [07:42] robert_ancell: this says to switch [07:43] robert_ancell: however, I hope that current upstream versions are better than what we have in karmic [07:43] the latter is absolutely useless [07:43] pitti, I just bought a scanner today and xsane is not great [07:44] robert_ancell: I know, but try gnome-scan [07:44] you can't even get the job done with it :) [07:44] robert_ancell: gscan2pdf is pretty good, but it was deemed too complex [07:44] pitti, I just tried it then and it does the scan but doesn't show anything :( [07:44] well, it has the workflow the wrong way around, no possibility to rotate or fix the brightness/gamma, etc. [08:02] pitti, yes, gscan2pdf does seem to work the best for me [08:02] thanks [08:12] hey seb128 [08:12] hey pitti! [08:12] hi seb128 [08:12] hi ajmitch, it has been a while, how are you? [08:13] I'm good :) [08:13] * ajmitch has been around a bit lately :) [08:13] yeah I noticed some uploads from you [08:13] I made some spare time :) [08:14] good! [08:14] hey ajmitch [08:16] * seb128 wonders why pitti apologized on the list [08:17] I start having enough of those mono wars there, they discussed the topic enough by now you could think [08:17] right, and I sort of threw the gauntlet again [08:17] "Never argue with an idiot. They bring you down to their level, then beat you with experience." [08:18] (completely ramdon quote which just came to my mind) [08:18] ;-) [08:18] random, too [08:18] I like it ;-) [08:18] * ajmitch understands your frustration :) [08:18] I just ignore those guys nowadays [08:18] especially the ones saying "I'm switching distro if you don't do what I want" [08:18] they sometimes provide amusement [08:19] current set of karmic packages are safe to run at the moment, aren't they? [08:19] seb128: Hi again. Just a heads up, if you use an NVIDIA card with X, the nvidia-180 drivers don't seem to build against 2.6.31 currently, so you may have to hold off your sound test with 2.6.31 if you use NVIDIA, at least till the dkms package is fixed. [08:19] ajmitch: really current current, with gdm fixed, yes [08:19] * ajmitch tends to upgrade by picking & choosing in synaptic, and saw the gdm issues yesterday [08:20] * TheMuso found that out today when wanting to test 2.6.31 with something else. [08:20] TheMuso, I'm using an intel card and sounds seem to be back after daily upgrades so maybe it was some other issues in the udev changes or something or resetting the pulseaudio, etc config made it work not sure [08:21] hey ajmitch, nice to see you [08:22] seb128: ok no problem then, good to hear you're back up and running with sound. [08:22] ...and on that note... [08:22] * TheMuso is outa here. [08:22] hi mvo [08:23] TheMuso: sleep well [08:23] seb128: \o/ [08:24] bye TheMuso [08:24] * pitti tosses some ♩ ♪ ♫ to seb128 [08:24] pitti, ;-) [08:24] * ajmitch shouldn't get sucked into reading debian-devel 'discussions' [08:26] what are they discussing now? [08:26] ia32-apt-get, or how not to do multiarch [08:26] I'm not reading debian-devel nowadays [08:26] probably wise [08:27] I was bored today & wondered what they were talking about in #debian-devel [08:37] seb128: btw, didn't MacSlow send a g-p-m patch recently? I thought you wanted to sponsor it? [08:37] is synaptic crashing for other people too in karmic when using the menu item to start it? [08:37] pitti, not sure, I don't do gpm, I sponsored notify-osd and g-s-d [08:37] ah, ok [08:37] I'm not subscribed to gpm bugs [08:37] will get it out from my mail archive then [08:38] seb128: btw, does new gdm still start for you? [08:38] it shouldn't? [08:38] seb128: it fails with "not the default WM" for me [08:38] $ cat /etc/X11/default-display-manager [08:38] gra [08:38] /usr/sbin/gdm [08:38] but the init script compares that to "/usr/sbin/gdm-binary" [08:38] yeah I changed that in my upload yesterday [08:38] I don't know how it was before this morning's upgrade, I'm afraid [08:38] I hate this thing [08:39] (it worked until yesterday) [08:39] because of the [08:39] SSD_ARG="--exec $DAEMON" [08:39] "restart" didn''t work otherwise [08:39] hi seb128 [08:39] hey robert_ancell [08:40] robert_ancell, had a good day today? [08:40] robert_ancell, when do you fly to desktop summit? [08:41] seb128, administration day today. I fly out tomorrow so I've bought a printer/scanner and did my paperwork I've been putting off. Wrote my GUADEC lightning talk too [08:41] what will you talk about? [08:43] "practical part-time refactoring". It's describes what I think is a good way to get into GNOME based on my experiences in gcalctool; [08:44] seb128: is synaptic crashing or gksu? [08:44] good morning everyone o/ [08:44] mvo, dunno, it doesn't open [08:45] seb128: could you please run "gksu id" in a term? [08:45] gksu synaptic in a g-t works correctly [08:45] Hey pitti, can you look at bug 195737 and bug 221698 again? They both have LP branches - should I add more information to upload requests than that? [08:45] Launchpad bug 195737 in compiz-fusion-plugins-main "Make window open animation the same as close animation (glide 2)" [Wishlist,In progress] https://launchpad.net/bugs/195737 [08:45] Launchpad bug 221698 in compiz "Cannot resize window taller than screen (inconsistent with metacity)" [Medium,In progress] https://launchpad.net/bugs/221698 [08:45] but not from gnome-panel apparently [08:45] seb128: how strange, anything in .xsession-errors ? [08:45] mvo, "sudo: sorry, a password is required to run sudo" in .xsession-errors [08:46] robert_ancell: ah, thanks for marking that other one as a dup; clear enough now [08:47] mvo, oh now it breaks in a g-t too [08:47] $ gksu synaptic [08:47] robert_ancell: fine for me now [08:47] sudo: sorry, a password is required to run sudo [08:47] $ gksu id [08:47] sudo: sorry, a password is required to run sudo [08:47] heh [08:47] mvo, ^ is gksu bog [08:47] seb128: I thought so :) [08:47] pitti, should I merge the two changes in BZR into one release? [08:47] seb128: let me have a look [08:47] robert_ancell: how else did you do it? [08:48] mvo, I get the issue on my laptop since yesterday and on my desktop now [08:48] pitti, there's two UNRELEASED entries in lp:~compiz/compiz/ubuntu [08:48] robert_ancell: right, please merge them into one [08:50] seb128: thanks, I can reproduce it now here [08:50] pitti, done [08:51] seb128, do you know if the DXE fusa work is going to be the upstream fusa solution? i.e. replacing the existing fusa releases? [08:52] robert_ancell, upstream has fusa in gdm now but dxteam wrote a new applet apparently rather than changing this one [08:53] seb128, So the one in http://ftp.acc.umu.se/pub/GNOME/sources/fast-user-switch-applet/ is deprecated? [08:53] NIH, or because the other one would have needed too much work? [08:53] robert_ancell, yes [08:53] seb128: yeah, it looks like gksu has changed the way it calls sudo in a strange way [08:54] * mvo wonders if its releated to the gtkbuilder patch from seb128 ;) [08:55] * seb128 slaps mvo [08:55] * mvo hides [08:55] is there a kind archive admin to accept new libgirepository1 binary package, part of gobject-introspection? :) [08:55] didrocks, doing [08:55] btw, is there a page listing all new binary built and waiting for acceptance (I saw that: https://edge.launchpad.net/ubuntu/+source/gobject-introspection/0.6.3-0ubuntu1/+build/1100440) [08:55] thanks seb128 ;) [08:57] seb128, are there any gnome-shell .debs around? [08:57] robert_ancell, gnagnagna [08:57] seb128, :P I have all day to think up questions for you [08:57] robert_ancell, working on it, cf didrocks gobject-introspection request ;-) [08:57] ooh gnome shell =O [08:58] I did upload gjs to the ubuntu-desktop ppa yesterday [08:58] cool [08:58] ie we are on it but there is a chain of depends to tackle [08:58] want to do some of the work? ;-) [08:58] didrocks: something like launchpad.net/ubuntu/karmic/+queue ? [08:58] didrocks, no, https://launchpad.net/ubuntu/karmic/+queue has the list of binaries [08:58] so, when libgirepository1 will be accepted, gir-repository will be able to build, and so clutter 0.9.4 will be built to \o/ [08:58] then, mutter is for me :) [08:58] seb128, I would say yes but I'm about to go travelling. I can have a look if I've got some spare time in GUADEC [08:59] seb128, ajmitch: ok, it's packages with NEW state, right? (so, New doesn't corresponds to "new uploads" [08:59] didrocks: need some NEWing? [08:59] robert_ancell, I said rick I would have it in a ppa before GUADEC so thanks [08:59] pitti, I'm on it [08:59] pitti: seb128 is on it :) [08:59] yeah, NEW binaries/source [09:00] didrocks, newed to universe [09:00] ok, thanks. didn't notice the icon difference at first glance :) [09:00] thanks seb128. Let's wait for the toolchain to be built so. [09:01] didrocks, you have local builds no? ;-) [09:01] that reminds me, I have a critical sync bug to get filed :) [09:02] * seb128 learnt a new trick [09:02] dput ppa:... [09:02] seb128: yes, I built them yesterday with that. But I'm currently at work and will be busy this morning ;) [09:02] didrocks, ok [09:02] seb128: you mean there's a way to do that without adding all of them to ~/.dputrc? [09:02] .dput.cf, rather [09:02] pitti, yes, "dput ppa:user/ppa_name .changes" [09:03] pitti, doesn't require any dput config [09:03] nice [09:03] for your favourite one it's still a bit too verbose, but good to know [09:03] ok. I know what I will test this evening :) [09:03] pitti, ie https://edge.launchpad.net/~ubuntu-desktop/+archive/ppa the first line [09:03] seb128: it works for team PPAs? [09:03] useful [09:03] * pitti grrrs at heisenbugs [09:03] ajmitch, yes [09:04] does anyone else have http://bugzilla.gnome.org/show_bug.cgi?id=585228 ? [09:04] Gnome bug 585228 in gnome-power-manager "immediately suspends on startup when lid is closed" [Major,Unconfirmed] [09:04] as soon as I start gnome-power-manager --verbose in the foreground, it doesn't happen any more [09:04] * ajmitch *still* hasn't upgraded his laptop from hardy :) [09:04] my laptop never suspended on lid close, is that a setting to configure? [09:04] seb128: it's the current g-p-m default [09:04] not for me [09:04] and what's annoying is that it happens in gdm [09:04] hm [09:05] ah, it coudl be due to dk-power d-bus activation [09:05] I close the lid all the team when I move between rooms [09:05] team -> time [09:05] seb128: I keep it closed when docked [09:05] and my laptop just lock screen [09:05] when docked I just suspend using the GNOME menu entry [09:06] seb128: that should be local gconf, though [09:06] since gdm has its own, it should happen there [09:06] I set it to "ignore" lid events on AC as well for my account [09:06] let me boot my laptop and try [09:06] hah [09:06] seb128: nevermind, I found it [09:06] ok [09:06] it's not a heisenbug, it's devkit-power sending a "lid event" when it's activated [09:07] even if the lid doesn't change [09:07] iz kit bog [09:07] ;-) [09:07] well, that, or g-p-m bug [09:07] but at least --verbose output says the right thing now [09:07] iz pitti bug in any case ;-) [09:07] seb128: right, I just wondered if I was going mad [09:13] bye all [09:29] seb128: I don't remember if I have to retry when the status is "Dependency wait" or just wait for a cron [09:31] didrocks, just wait [09:32] thanks [09:33] didrocks, the publisher take a while, those should be available in 15 minutes or so [09:34] pitti, wb [09:34] my server is offline, seems that the hosting DC lost the interweb tube [09:35] again? [09:35] that's twice in a week now [09:35] last time they did some hardware maintenance [09:36] the ppa publishing is fast nowadays [09:36] pitti, ok, the gdm default manager things is fixed in the ppa now [09:36] seb128: rock, thanks [09:36] binaries are published and available by normal apt upgrade [09:36] seb128: publishing> indeed, it's a breeze [09:37] hum [09:37] there is another annoying upgrade issue there [09:38] the fusa applet doesn't work after upgrade [09:42] seb128: I remembered that when I tried KDM, fusa applet was crashing when launching GNOME [09:43] didrocks - thats a known issue. quite a few users complain about that === asac_ is now known as asac === mdz_ is now known as mdz [10:22] re [10:22] didrocks, gir built now [10:31] lut huats [10:34] hello seb128 ! [10:35] how are you ? [10:37] huats, good, and you? [10:38] good too ! [10:38] the weather is getting too warm for my taste though [10:38] seb128: please don't mention the weather [10:38] i can't stand the warm :) [10:38] (well that kind of warm...) [10:39] and I am not going to the GUADEC... imagine what you are about to live :P [10:39] should be fine [10:39] grand canary is an island the forecast says 24°C with some wind there [10:40] it's cooler than the continual weather there [10:42] and theer is the sea... which helps A LOT ! [10:43] indeed [10:43] and ice cream too in case it's hot ;-) [10:44] :) [10:44] LOL [11:12] pitti: we're having a translations meeting tomorrow, and I'd like to tell translators what's currently happening around translations. I've been following bug #123020 and I think this would be a good thing to comment. Can I explain them the following: [11:12] Launchpad bug 123020 in soyuz "support shipping verbatim files in the exported tarballs" [High,In progress] https://launchpad.net/bugs/123020 [11:12] 1) There's work in progress in getting the GNOME help files in language packs [11:12] 2) This will _probably_ be done for Karmic [11:12] 3) At this point this will only be a LiveCD space-saving feature, since the xml to PO (import) and PO to xml (langpack export) conversion will still have to be done manually. [11:13] dpm: hi [11:14] dpm: s/LiveCD/CD/ (applies to alternate, too), otherwise correct [11:14] dpm: we will only copy the verbatim bits, no xml <-> po conversion for now [11:14] that requires proper rosetta support [11:15] brb, rebooting [11:15] pitti: hi :), thanks [11:17] pitti: ok, yes, that's what I meant with point 3), I should have probably formulated it better. I'll add the item to the agenda, then [11:18] doh, even pitti start using quilt now [11:22] quilt > * [11:22] *grumpf* I I finally found the problem with gksu, how anoying [11:23] mvo, you rock! [11:23] mvo, tell me it's not the gtkbuilder changes ;-) [11:23] seb128: well, just so that Michael Biebl doesn't kill me after my first commits to pkg-utopia :) [11:23] seb128: but compared to the pain of git, quilt is just worth a shallow "oh" [11:24] btw speaking about git how do you add a file and make it been listed in git diff too? [11:24] seb128: the new forkpty() behaviour and sudo tty_tickets do not work togetether [11:24] I shouldn't have to use patches in the first place [11:24] git add newfile; git diff lists nothing [11:24] but with git-buildpackage at least there seems to be no way to cherrypick patches from trunk properly [11:24] seb128: ah [11:24] seb128: "git diff" does not what you think [11:24] git diff HEAD^ ? [11:24] it shows the diff between the current index and the source [11:24] crevette: oh, so that's it? [11:25] * crevette is always lost with git diff [11:25] I always use "git commit -v" and cancel [11:26] pitti, I admit I didn't look deeply in the doc, as the man page is quite complicated, (you must know a lot of details of git inner to understand it) [11:26] git diff and git diff --cached [11:26] git diff shows the difference between your working tree and the index, git diff --cached shows the difference between HEAD and the index [11:26] hyperair, can I have both in the same command? [11:26] crevette: hah, tell me about it [11:26] seb128: git diff HEAD [11:27] i think [11:27] git is teh pain [11:27] git is really the suck to use [11:27] it's a shame that GNOME picked this one [11:27] better than bzr =p [11:27] I still didn't learn (and don't want to) all those concepts of heads, index, origins, checkouts, intra-tree branches, and how the heck to get a branch ssh'ed over there (i. e. bzr push) [11:27] bzr add && bzr diff does what you expect though [11:27] because bzr doesn't have a staging area [11:28] the fun thing is that even the folks who have used git for a long time don't know how to get git push working :) [11:28] I don't care about a staging area [11:28] pitti, I just had a basic usage as I wrote some patch for GNOME, but nothing fancu [11:28] fancy [11:28] hyperair: bzr shelve [11:28] I want to checkout, do changes and see my changes [11:28] pitti: isn't that the same as git stash? [11:28] anyway, we had enough of them [11:28] (flamewars, I mean) [11:28] hyperair: quite possibly [11:28] yeah so it's not the staging area after all =\ [11:29] but the index concept is so utterly confusing, and not really necessary either [11:29] seems git users are rather interested in how the tool can do tricky things than in getting work done easily [11:29] i don't like that bzr has so goddamn many repository formats, all of which are incompatible with one another [11:29] git stash is for on-going patch that break a rebase from I understood [11:29] i've had a few repositories turn unmergeable because of that [11:29] stupid bzr [11:29] crevette: no, git stash is to stash away your changes and return to a clean tree to do something else. [11:29] git stash put your change is a temporary place, and you can re-apply with a git pop [11:30] git stash apply [11:30] so as bzr shelve [11:31] bzr doesn't have the staging area [11:31] either way, i don't really use it much. i commit with git ci -av [11:32] I'm not sure what a staging area is and why I would want one [11:32] hyperair: no staging area> that's a feature :) [11:32] I usually checkout, commit commit commit, push somewhere for review [11:32] you can pretty much do that with git too [11:32] i like how git repositories are pretty much standardized [11:32] what the vcs does is implementation details I don't want to care about [11:32] hyperair: (honest question) do you know how to push a branch to my server, in a way that works? [11:33] instead of bzr godknows what knit rich root poor root whatever formats [11:33] pitti: push a bzr branch or a git branch? [11:33] hyperair: the best answer for that so far was "git prune; git gc; rsync -r .git myserver:public_html/git/stuff" [11:33] hyperair, git, bzr is easy, bzr push location ;-) [11:33] hyperair: with git (with bzr, it's just "bzr push bzr+ssh://myserver/dir") [11:33] hyperair: but with that prune/gc/rsync, I still can't push to that branch again [11:33] pitti: i think you need to initialize a bare repository first, even in the case of bzr. [11:34] hyperair: bare repo> no, tried that, doesn't work [11:34] pitti: prune and gc are just to reduce what you need to transfer [11:34] hyperair: (and that's not necessary with bzr; it just works) [11:34] yeah, it just transfers evry damn thing [11:34] anyway, I read docs and tried several recipes after 1.5 hours, then I just gave up and used the usual patch/git add/git commit pain [11:35] and mailed patches around [11:35] Pidgin 2.5.8 [11:35] pitti: chances are you want to push more than once, so you just add the remote first: git remote add origin myserver:public_html/git/stuff.git, then you do git push --all origin [11:35] (libpurple 2.5.8) [11:35] \o/ [11:36] pitti: the pain is only when you don't know how to use it. the learning curve is steep as hell, but once you understand it, it's hard to look back. [11:36] hyperair: right, that's what I meant; it's a tool which is written for people who want to learn it inside out, not for the casual developer who wants to work on source :( [11:36] instead of git [11:36] I still not get why people fancy a tool hard to learn when you have much easier alternative doing all the common tasks you need as well, but shrug [11:37] seb128: because the tool is awesome once you understand it. [11:37] seb128: because once you do know how to use it, it is that much better than other, easier tools. [11:37] and git is perhaps the tool for hardcore developers? not for casual hacker [11:37] I don't care about how the tool is awesome, I want to checkout, commit commit commit, push somewhere [11:37] the only real advantage that I see is speed [11:37] same question as why do people use vim instead of nano [11:37] and any modern vcs allows that [11:37] but for almost all projects that doesn't matter [11:37] pitti: size. [11:38] hyperair: I don't care [11:38] git is huge, sure, but even that I don't mind [11:38] (git trees, I mean) [11:38] pitti: i recovered somewhere around 1G by shoving every thing into git from bzr. [11:38] hyperair, I don't see how it can be better than bzr for checkout, commit, push which is all I need [11:38] pitti: granted, i had a bunch of tarballs lying around that got shoved into git as well [11:38] my projects are in the 50 to 500 KB range, I don't care at all about speed or size [11:39] seb128: if that's all you need, and are sure you don't need any more, then no, it's no better than bzr. [11:39] hyperair, and all what 99% of people need with easy merging and some other things which are available in any vcs [11:39] obviously git is very good at handling kernel trees, and bzr sucks for that [11:39] seb128: i found git merging much smoother than bzr's. [11:39] but 95% of the projects out there aren't kernel sized [11:39] seb128: at the very least, i didn't have to worry about repository incompatibilities [11:39] hm [11:39] hyperair, I usually bzr merge, review, commit [11:39] last time I merged, it utterly failed [11:39] I ended up having to format-patch there, and git patch/git add/git commit (times 10) [11:40] hyperair, you seem to focus on this format issue, I never got bitten by that though, no reason the format should change in a set up repository [11:40] * pitti STFU now [11:40] i got bitten MANY times. [11:40] so many that i switched to git [11:40] hyperair: let's agree to disagree [11:40] indeed [11:40] you won't ever convince me, and apparently I won't convince you :) [11:40] good the poppler guys merged my patch to use gtkbuilder [11:41] one less libglade depends for karmic ;-) [11:41] hah, Richard ack'ed my devkit-power patch to fix the "immediately suspends" issue \o/ [11:41] pitti: a year or so ago, you would have. rather, i read the docs and was convinced to use bzr. then the format issue bit my ass so many times that i switched to git, and am happy to say i'm not going back. [11:41] pitti: pristine-tar rocks, by the way. [11:42] hyperair: pristine-tar> *nod* (that exists for both systems) [11:43] pitti: yeah it does. but where does it store the .delta files for bzr? [11:43] I don't use it for my projects, but pkg-utopia does, and it's quite an interesting idea [11:43] pitti: in git, it just dumps it into the tree under the "pristine-tar" branch. [11:43] hyperair: I don't know (and I don't care :) ) [11:43] it's pretty awesome for package maintainenace i must say =p [11:44] rather than 10s of tarballs, you just have a VCS and pristine-tar deltas [11:44] right [11:47] * hyperair wonders where bdefreese has gone. [11:47] nautilus-share needs a sponsor into debian =\ [11:48] ask ajmitch! [11:48] he's a DD? [11:48] when will debian accept source uploads... [11:48] yes [11:49] doesn't it? [11:49] * Laney plays with the new bzr branches [11:49] hyperair: no, you always have to upload a set of binaries alongside [11:49] what happens if you upload _source.changes? [11:49] * hyperair patiently waits for git-bzr to enter the archives [11:49] will dak reject it? [11:50] hmm that doesn't sound nice =\ [11:50] I would sponsor some uploads otherwise [11:50] the binaries get thrown away anyway, dont' they? [11:50] no [11:50] but I don't have a current debian unstable install to build binaries [11:50] seb128: pbuilder! [11:50] no the binaries you upload are the one available on the mirrors [11:51] pbuilder-dist sid create [11:51] which means that if you have experimental packages, etc you can screw builds easily [11:51] I know how to use pbuilder [11:51] :) [11:51] the issue is how much daily download you have to do [11:52] i don't have much =\ [11:52] ie I'm rsyncing daily karmic CD now and it will saturate my download for 2 hours still [11:52] oh [11:52] heh [11:52] then I dist-upgrade karmic and it's one extra hour [11:52] and it's lag so much that I can't triage bugs [11:52] I don't fancy putting my internet in lag mode for an another hour only to update a sid pbuilder [11:52] * hyperair thinks living on campus with a server on LAN is awesome =p [11:53] I would be happy to test nautilus-share on karmic and do a source upload to debian though [11:53] but since that's not possible you will have to find another sponsor [11:53] so yeah, does dak actually reject it? [11:53] source uploads are not accepted no [11:54] what's the point of taking source updates anyway =\ [11:54] what if i build nautilus-share for you? [11:55] sources update means you have less to upload and you are sure that the build is made on a clean install [11:55] I could build on karmic and upload to debian and the binary would depends on gtk 2.17 which is not in debian [11:55] mmhmm [11:55] what is the rationale? buildd time? [11:56] no they basically argue that it forces maintainer to test build before uploading [11:56] ie you can't upload something which obvious ftbfs [11:57] it's sort of a wrong technical solution to a social issue [11:57] nothing like a bit of trust eh [11:57] heh yeah [11:57] it does solve some edge problems though i.e. bootstrapping [11:57] any lunch time [11:57] bbl [11:57] k [11:57] Laney, not really since you upload binaries for one arch only [11:57] it will still need to build on the buildds for other architectures [11:58] arch=all stuff =p [11:58] right [11:58] yeah [11:58] anyway, pidgin diff is up [11:58] lunch and travelling to the office, bbl [11:59] Laney, thanks, I will review that after lunch [11:59] bbl [12:00] Laney: by the way, you might be interested to know that the geany plugins have merged and will soon release (i just have to testbuild and tag the release) [12:01] didrocks, clutter-0.9 building now [12:01] seb128: thanks for noticing :-) I will have some time now to dedicate to mutter. [12:02] seb128: ugh, what downstream do you have? [12:04] hyperair: merged how? [12:04] Laney: geany-plugins-0.17.tar.gz [12:04] pitti, my dsl? 1024,256 [12:04] seb128: ugh, poor you [12:05] hyperair: oh, so all of the other source packages now go away? [12:05] mmhmm [12:05] fun [12:05] not all [12:05] geanyprj didn't get merged [12:05] not all of them joined [12:05] but most did [12:05] i see [12:05] any progress on debian? [12:05] pitti, well it's ok to do daily karmic updates, but not for updating karmic, debian, jaunty, iso download, etc [12:06] Laney: got sponsor, but we decided to wait for geany-plugins to release. [12:06] \o/ [12:06] =) [12:18] seb128: new libgksu should fix the problem, pleae le tme know (but I'm off for lunch now) [12:23] what is this super-l email about? [12:27] I thought it was a spam [12:27] :) [12:28] I have no idea what the original problem is [12:28] is it just a nutjob? [12:29] i thought the same thing [12:29] it seems a bit wierd [12:41] mvo, cool, I will test when it's available [12:41] Laney, what email? [12:45] seb128 - it's another mail on u-d-d ;) [12:46] chrisccoulson, weird I didn't receive it, maybe it has been tagged as spam on the way there [12:47] quite possibly. i thought it was spam when i first opened it but then couldnt work out whether the poster was being serious or not [12:47] and i was still half-asleep when i read it! [13:37] seb128: can't I use 'debian/tmp/*' in *.install files? [13:37] rodrigo_, you can, why? [13:37] seb128: it fails with no files found, but 'make install' installed them [13:37] seb128: http://launchpadlibrarian.net/28596352/buildlog_ubuntu-karmic-lpia.evolution-couchdb_0.1.1-7_FAILEDTOBUILD.txt.gz [13:37] where is the source package? [13:38] seb128: in https://launchpad.net/~rodrigo-moya/+archive/ppa [13:39] rodrigo_, you have only one binary, it does a make install directory to evolution-couchdb there is no .install [13:39] rodrigo_, the .install are required when you have several binaries to say what goes where [13:40] ah, ok, so with no .install file it would just include everything it installed? [13:40] rodrigo_, you might override that in the rules and say to install to tmp and move things if you really want though [13:40] rodrigo_, yes when there is only one deb built the default is to ship everything [13:40] ok, cool [13:40] ie make install to debian/binary [13:42] * rodrigo_ feels he's almost there :) [13:43] ;-) [13:45] the email was to ubuntu-desktop [13:46] oh, I got this one and marked it as read without reading it apparently [13:46] the title probably inspired me in my morning cleaning ;-) [13:47] what's it about? anything? [13:48] it seems long and boring I'm not really interested to read it to say ;-) [13:49] chrisccoulson, there is a new gnome-session, interested by doing the update? [14:08] seb128 - yeah, i can do the gnome-session update [14:08] chrisccoulson, thanks [14:08] no problem [14:28] didrocks: howdy [14:28] Mr. Spencer! [14:28] hi pitti [14:29] hey rickspencer3 [14:29] hi seb128 [14:29] rickspencer3, when do you start traveling to desktop summit? [14:30] seb128: late morning [14:30] mvo, just tested your libgksu update that fixes the issue! [14:30] hey rickspencer3 [14:30] after I finish slurping my coffee, I'll pack, and then pull some code [14:30] then I'm off [14:30] rickspencer3, is quickly all ready for your talk? [14:30] kenvandine: the coding part is [14:31] it would be cool if I could demo packaging and/or releasing, but that's not too important [14:32] rickspencer3: hey rick! [14:33] didrocks: I pimped out https://code.edge.launchpad.net/quickly a little bit [14:33] rickspencer3: yes, with a new logo :D [14:33] I'm going to pull in the next hour so, do you have code to push before I leave? [14:34] rickspencer3: I wanted to try to finish quickly release tonight if "mutter" doesn't give hard time :) [14:34] didrocks: I put the svg file in with the ubuntu-project template, so feel free to make the logo better [14:34] rickspencer3: will you be able to pull when arriving to GUADEC? [14:34] didrocks: yes, for sure I'll pull every day :) [14:35] rickspencer3: I will try to not push too many thing so that you don't have the "demo effect" :) [14:35] just finish nice stuff in quickly release [14:35] seb128: excellent, I send a mail to kov discussing the issue, I think its a design problem with the latest code, but I'm curious to hear his opinion === maxb__ is now known as maxb [14:37] rickspencer3: if you want to get ride off all the config, you can remove ~/.quickly-data (it's where cache and crendentials are stored) [14:37] didrocks: go ahead and push! [14:37] I can roll back if there's a problem [14:37] mvo, ok cool [14:38] rickspencer3: ok. Not at home currently. I will do this tonight. So, when you will be already in your plane :) [14:38] didrocks: right ... I'll work on the tutorial code and the tutorial, and on the template ... [14:38] so we shouldn't have any conflicts to merge [14:39] * rickspencer3 hates merging conflicts [14:39] rickspencer3: yes, I saw it gave you hard time. Next time, let me merge it (I have a bzr branch in my server) [14:40] didrocks: :) === SiDi_ is now known as SiDi [14:58] geser: re bug #384144 if you can rperoduce that easily, could you check if apport gets a backtrace ? [14:58] Launchpad bug 384144 in ubuntuone-storage-protocol "debian/copyright should mention GPL files" [Undecided,New] https://launchpad.net/bugs/384144 [15:01] kenvandine: ping [15:03] Mark__T, pong [15:03] Mark__T, what's up man? [15:03] will karmic have indicator-applet 0.2? [15:03] yes [15:04] Okay, SiDi sort of asked me on a different channel, so I thought, I make sure [15:04] :) [15:04] Mark__T, how current is the snapshot in foresight? [15:05] kenvandine: latest revision [15:05] ok [15:05] 309 [15:10] mvo: do you mean bug #385144? will try to get apport catch something [15:10] Launchpad bug 385144 in apt "apt-get dies with "E: Method http has died unexpectedly!"" [Undecided,New] https://launchpad.net/bugs/385144 [15:11] kenvandine: will it have libindicate in a non-gnome-dependant package ? ^.^ [15:16] geser: that is much appreciated - I tried it with squid on a tmpfs pbuilder here and I get a random hang for some seconds, but no crash [15:16] yet [15:17] geser: what kind of download rates you get with the cacher? [15:18] pitti: hey. did you get a chance to look at ubuntuone-client ? [15:19] mvo: apt shows in the summary up to nearly 60 MB/s for downloading the build-depends for some packages I tried to build [15:19] rickspencer3: you leave today for GC right? [15:19] jcastro: yes [15:19] heading to the airport in a couple of hours, packing now [15:19] geser: I suspect this is key to reproduce, I get only ~30 so far [15:20] rickspencer3: me too, what time do you get there locally? I get there 2pm Thursday local time [15:21] jcastro: are you flying from Madrid? [15:21] dobey: on my list; I started reviewing last night, but then I just fell asleep [15:21] rickspencer3: yeah, AA6240 [15:21] geser: what is the package you get it with? or does it happen with multiple ones? [15:21] dobey: it needs a couple of fixes such as the new python-ubuntuone-storage-backend dependency, and some other stuff I forgot (will do detailled review later) [15:23] jcastro: I'm getting in at 2:45 on spain air [15:23] maybe I'll see you at the airport in Madrid [15:23] ok [15:23] rickspencer3: it's only 45 minutes, I'll just wait for you at the GC airport [15:23] I have all the maps and stuff [15:24] pitti: doh. i see i forgot a comma in one place [15:25] dobey: there were some other lintian things as well [15:26] dobey: do you use lintian? it's quite helpful [15:26] it is not clear how to, and debuild doesn't seem to run it i guess [15:26] it does [15:26] you just need to install it [15:26] it is installed [15:27] kenvandine: mind if i pm you ? [15:27] hm, it's run by default after debuild -S or debuild -b [15:27] SiDi, sure [15:27] dobey, i saw it when i built it :) [15:27] hrmm, then it must be very sparse in jaunty or something, because i ran debuild -S -sa to build the source package [15:27] because i fixed the few things that were complained about in its output [15:27] dobey: source package might be okay; the binary checks are (and can) only be run if you build binaries [15:28] well i did a binary build as well to make sure i had the ${source:Version} bit correct because my memory was vague on it [15:28] jcastro: thanks [15:28] that would be really great [15:28] I'll try to meet up with you in Madrid too [15:28] and it didn't complain about anything extra that seemed to warrant fixing (Original-Maintainer and Python-Version or something was all it mentioned) [15:32] seb128: another question :) to add a patch to a package, I just need to copy it to debian/patches, right? [15:37] rodrigo_ - it depends what patch system the package uses [15:37] chrisccoulson: hmm, how do I find out? [15:38] it's got no patches right now, or do you mean something else? [15:39] you could take a look in debian/rules. if it's got no patches already, then you might need to add a patch system in debian/rules [15:42] chrisccoulson: how do I do that? (could you point me to some docs?) [15:43] * seb128 kicks his isp today [15:43] oh, it already has: include /usr/share/cdbs/1/rules/simple-patchsys.mk [15:43] is that ok? [15:43] rodrigo_, add a "include /usr/share/cdbs/1/rules/simple-patchsys.mk" in rules [15:43] rodrigo_, then you can copy a patch in the patches directory and get it used automagically [15:43] ah ok [15:43] cool [15:44] rickspencer3: https://blueprints.edge.launchpad.net/ubuntu/+spec/desktop-karmic-automagic-python-build-system :) [15:44] pitti: wow! [15:44] lots of "DONE"! [15:44] python-mkdebian now generates a working and reasonably policy compliant debian package from apport trunk [15:45] all imports are correctly converted to Dependencies, etc. [15:45] I can't wait to try it on the ubuntu-project [15:45] wow! [15:45] that is so cool [15:45] I think we can use this for quickly itself! [15:45] some stuff is still missing [15:45] such as updating existing debian/control, test cases, proposing distutils extra base stuff as a PEP, etc. [15:45] but it should be enough to get you going [15:46] right [15:46] (test cases for debianization, that is; DistUtilsExtra.auto has full test coverage) [15:46] I should be able to create $quickly package fairly easily now [15:46] shoudl be on archive.u.c. in 75 minutes [15:47] or, if you want to try it now, lp:~python-distutils-extra-hackers/python-distutils-extra/debian [15:47] mvo: it happened with different ones, currently I'm reproducing it with camlimages [15:47] mvo: how do I get apport to pick something up? as apt doesn't really crash but exits with the message and return code 100 [15:48] eww, I suck [15:49] geser: I thought it might crash, but if it does not I add some better error output and do a upload in some minutes [15:49] pitti: I'll pull, as I'm bringing a Jaunty machine to Gran Canaria (my eee) [15:50] mvo: http://pastebin.ubuntu.com/207565/ that's what I get when I try to download the build-depends for camlimages [15:52] geser: thanks, I try this now [16:11] seb128 rickspencer3: We're getting no movement on this week's paper cuts, when you get a moment, can you take a look and let me know if any of them look particularly suited for any individual contributors? I can ping them (or you can) to try to get some of them looked at. https://bugs.edge.launchpad.net/hundredpapercuts/+milestone/round-2 [16:26] djsiegel, I will have a look now but this week is not the best week to get work on those since a good part of the desktop people are travelling to the desktop summit [16:26] seb128: right, including me [16:26] :) [16:29] djsiegel, it would be nice it you could update the description with the design suggestion [16:29] djsiegel, it's not always easy to find the recommendation in the list of comments [16:29] seb128: ok, I was wondering about that, if it was good or bad bug etiquette [16:29] djsiegel, ie bug #23293 it's not clear what should be fone [16:29] Launchpad bug 23293 in nautilus "Nautilus windows should close on unmount" [Unknown,Confirmed] https://launchpad.net/bugs/23293 [16:29] done [16:30] mvo: I get "E: ::ReadMessages returned false - pkgAcquire::Worker::ReadMessages (2 No such file or directory)" with your new apt [16:32] geser: excellent, thanks a lot! could you please add the debug string I put in the report too ? (https://bugs.launchpad.net/ubuntu/+source/apt/+bug/385144/comments/4) ? [16:32] Ubuntu bug 385144 in apt "apt-get dies with "E: Method http has died unexpectedly!"" [Undecided,Confirmed] [16:41] geser: I also uploaded another apt with different debug output now [16:53] geser: hm, you said the method exists with exit status 100 ? [16:55] mvo: yes, echo $? returns 100 [17:04] seb128 - would you mind adding some comments to my MOTU application when you have some spare time? :) [17:05] mvo: I've attached the apt output (with debugging) to the bug, trying ubuntu1.2 now [17:08] chrisccoulson, that's on a wiki page right? [17:09] chrisccoulson: me too, me too! [17:11] pitti: The decision for now is to use glide-2 for open and close (https://bugs.edge.launchpad.net/ubuntu/+source/compiz/+bug/195737) [17:11] Ubuntu bug 195737 in compiz-fusion-plugins-main "Make window open animation the same as close animation (glide 2)" [Wishlist,In progress] [17:11] djsiegel: I saw, thanks; I'll sponsor that soon [17:11] cool, thanks [17:11] Not sure why those two bug reports got swizzled into one [17:11] :) [17:12] pitti, seb128 - the application is here: https://wiki.ubuntu.com/ChrisCoulson/MOTUApplication [17:12] you mean "into two"? [17:12] chrisccoulson, ok [17:12] thanks:) [17:14] mvo: "E: ReadMessages: read return 0" with apt 0.7.21ubuntu1.2 [17:27] djsiegel, bug #372132 needs design recommendations [17:27] Launchpad bug 372132 in hundredpapercuts ""Create Document" Templates difficult to use" [Undecided,Confirmed] https://launchpad.net/bugs/372132 [17:27] djsiegel, bug #387796 too [17:27] Launchpad bug 387796 in hundredpapercuts "Display icons in the "open with ..." list " [Undecided,In progress] https://launchpad.net/bugs/387796 [17:29] djsiegel, bug #390362 added to review queue it's probably for asac [17:29] Launchpad bug 390362 in hundredpapercuts "entering wep key to NM requires tabbing/clicking" [Undecided,In progress] https://launchpad.net/bugs/390362 [17:29] djsiegel, bug #146918 needs design recommendations [17:29] Launchpad bug 146918 in update-notifier "Poor descriptions for some applications in Startup Programs window" [Low,Confirmed] https://launchpad.net/bugs/146918 [17:29] djsiegel, bug #392292 needs design recommendations [17:29] Launchpad bug 392292 in hundredpapercuts "Tooltip of the search button[nautilus] is misleading" [Undecided,In progress] https://launchpad.net/bugs/392292 [17:29] seb128: yes, I know [17:30] djsiegel, well it seems to me that the design team should have made recommendations earlier than that if you want people to tackle those issues [17:30] djsiegel, bugs are not being worked because there is no clear instructions on what need to be done on most of the bugs for this round [17:31] seb128: that's true, I know. I don't think the design team is clear on their role. The idea isn't that all design on every bug would be done by the design team. [17:31] seb128: before there was a canonical design team, weren't there people working on usability bugs? [17:32] djsiegel, well, "descriptions are not good enough" requires an english speaking to suggest better wording for example [17:32] djsiegel, not really, which is part of why those bugs have not been fixed yet ... [17:32] speaking -> speaker [17:34] seb128: I'm afraid 100 paper cuts will not be finished if people are waiting for one of the three or four of us on the design team looking at paper cuts to adjudicate on each one. [17:34] djsiegel, ok so they will not be finished, fair enough [17:34] seb128: we are putting the spotlight on these and helping user test the hard ones, but the operating assumption should not be that they are blocking until the design team weighs in. [17:35] djsiegel, I'm not an english speaker and I don't know what would be better wording for those labels that what we have now for example [17:35] Many of the bugs fixed so far have been fixed without design team suggestion [17:35] seb128: ok, don't propose a solution for that bug then [17:36] in fact, on bugs where someone from the design team has weighed in, it has slowed bug progress [17:36] djsiegel, those I listed before are blocked on somebody having a clue about design making a recommendation [17:36] well "Create Document" Templates difficult to use" [17:36] ok, well, I am working on one right now [17:36] take that one for example [17:36] and I emailed the team to get moving [17:36] great, why is it difficult? [17:36] if it's open for 5 years that's a chance that the hackers don't find it difficult to use [17:37] and will not make change if you don't suggest an easier way [17:37] right, well, most of the time we agree with the proposed solution in the bug description/title [17:37] if we need to restate that and say "this is the suggestion," we can do that [17:38] same for bug #23293 [17:38] Launchpad bug 23293 in nautilus "Nautilus windows should close on unmount" [Unknown,Confirmed] https://launchpad.net/bugs/23293 [17:38] what else should it be doing? [17:38] there are no perfect solutions [17:38] ups, wrong bug number [17:38] we are just looking for progress on some [17:38] well there is some progresses [17:39] the round-2 list has 1 fix commited, one patch waiting for upstream review and a bunch of in progress bugs [17:39] djsiegel, bug #23293 I disagree with for example, it has been changed to do that because we were getting bugs from user saying that nautilus was crashing [17:39] Launchpad bug 23293 in nautilus "Nautilus windows should close on unmount" [Unknown,Confirmed] https://launchpad.net/bugs/23293 [17:40] djsiegel, for users apparently "closing without action" == crashing [17:40] seb128: reasonable, we could user test, but there is action [17:40] ejecting [17:40] well apparently users don't do the connection [17:41] I can look for old bugs about that if that's of any use [17:46] geser: many thanks \o/ [17:46] awe: can you look at #390362 please [17:47] its a papercut bug. not sure what we want there or not. [17:47] asac: sure [17:48] I actually mentioned it to rick last night [17:48] awe: if you think it makes sense in general, we might want to apply it for now [17:48] it seems like a good fix, but I haven't looked at the patch closely yet [17:48] even if its useless in future [17:48] I'll take a look this afternoon. [17:49] thx [17:49] djsiegel, ok, I've added ubuntu task for bugs which didn't have so the maintainers get notified and added some comments [17:49] np [17:52] seb128: awesome, and I added a design suggestion: http://launchpadlibrarian.net/28602360/dialog.png [17:52] cool [17:53] good night everyone, time for dinner and Taekwondo [17:53] good night pitti [17:54] 'night pitti [18:39] geser: I uploaded yet another apt with debug output :) === pochu_ is now known as pochu [23:10] asac, there? [23:16] seb128: yes [23:17] asac, do you plan to ship libmozjs in /usr/lib for karmic? [23:17] asac, I'm fighting gnome bug #573413 [23:17] thats the question [23:17] Gnome bug 573413 in general "libmozjs linking broken on Ubuntu" [Normal,Unconfirmed] http://bugzilla.gnome.org/show_bug.cgi?id=573413 [23:17] asac, ie trying to get gnome-shell in a ppa before GUADEC as asked by rick [23:18] gnagnagna, dan's patch doesn't do the trick [23:18] seb128: for now use LD_LIBRARY_PATH=/usr/lib/xulrunner-`xulrunner -gre-version`:$LD_LIBRARY_PATH [23:19] asac, where do I put that? [23:19] to get the package going [23:19] seb128: you need some start script ;) [23:19] asac, it ftbfs right now, it's not runtime [23:20] the build call g-ir-scanner to build some gobject introspection code [23:20] and that's the think which breaks [23:20] let me check the patch [23:21] seb128: how does the biuld fail? [23:21] without looking context i would say: 1. drop the configure part of the patch. 2. replace + mozjs_libdir = '@MOZJS_LIBDIR@' [23:21] "(mutter:6014): mutter-WARNING **: Could not load library [/builddir/src/libgnome-shell.la (libmozjs.so: cannot open shared object file: No such file or directory)] [23:21] (mutter:6014): mutter-CRITICAL **: failed to load plugins [23:21] Invalid GType: 'big_box_get_type' [23:21] failed to dump: Failed to find symbol 'big_box_get_type' [23:21] Command '['mutter', '--mutter-plugins=/builddir/src/libgnome-shell.la', '--introspect-dump=/tmp/tmp-introspectGeuNxi/types.txt,/tmp/tmp-introspectGeuNxi/dump.xml']' returned non-zero exit status 1" [23:22] with mozjs_libdir = /usr/lib/xulrunner-`xulrunner -gre-version` [23:22] and dont use the @...@ thing detected at build time [23:22] $ grep mozjs_libdir * [23:22] gnome-shell: mozjs_libdir = '/usr/lib/xulrunner-1.9.0.11' [23:22] seb128: yes. but thats wrong. it should always be evaluated dynamically [23:22] well that will not make a different in this case [23:22] I've the lib in '/usr/lib/xulrunner-1.9.0.11 [23:23] yes [23:23] thats not the build problem [23:23] but the runtime problem you will end up [23:23] I'm wondering if I can add a -rpath somewhere [23:23] i need to see the link linne of libgnome-shell.la [23:24] seb128: i think setting LD_LIBRARY_PATH during build correctly might also help [23:24] if its works its better than -rpath because its just for the build and doesnt influence runtime behaviour later on [23:25] asac, http://pastebin.ubuntu.com/207770/ [23:26] seb128: and the content of libgnome-shell.la? [23:26] asac, http://pastebin.ubuntu.com/207772/ [23:27] seb128: i guess you have the /usr/lib/xulrunner-devel-1.9.0.11/ ? [23:27] export LD_LIBRARY_PATH=$LD_LIBRARY_PATH:/usr/lib/xulrunner-1.9.0.11 [23:27] make [23:27] that works [23:27] yes [23:27] thats the idea [23:28] and then fix the gnome-shell like i said [23:28] and you should be fine [23:28] I'm just not sure where to plug the LD_LIBRARY_PATH during the build [23:28] I will figure thanks [23:28] seb128: debian/rules [23:28] just set it there [23:28] how? [23:28] export .... ? [23:28] export LD_LIBRARY_PATH=... [23:28] yes [23:28] I tried but it didn't like that [23:28] i think you can even just use export LD_LIBRARY_PATH="/usr/lib/xulrunner-$(shell xulrunner --gre-version)" [23:29] hmm [23:29] cdbs? [23:29] lot of [23:29] ERROR: ld.so: object 'libfakeroot-sysv.so' from LD_PRELOAD cannot be preloaded: ignored. [23:29] maybe there is a special hook [23:29] yes [23:29] seb128: ah [23:29] yeah you need to use =$LD_LIBRARY_PATH:... [23:29] e.g. keep the original path [23:29] thats where fakeroot is [23:29] I used [23:29] +export LD_LIBRARY_PATH=$LD_LIBRARY_PATH:/usr/lib/xulrunner-1.9.0.11 [23:30] seb128: hmm. doesnt make evaluate that up front? [23:30] try $$LD_LIBRARY_PATH [23:30] no warning, let's see if that makes what I want too [23:31] hum, I got those warnings later during the build [23:32] otherwise that seems to work [23:32] ie it builds [23:32] ok [23:32] it just stop on "install: cannot change owner and permissions" due to the preload issue [23:32] hmm [23:32] lets check cdbs [23:33] nothing [23:33] but that might not be due to the rules change [23:33] Tidy-1.0.typelib: libtidy-1.0.la Tidy-1.0.gir [23:33] LD_LIBRARY_PATH=$${LD_LIBRARY_PATH:+$$LD_LIBRARY_PATH:}. g-ir-compiler Tidy-1.0.gir -o $@ [23:33] CLEANFILES += Tidy-1.0.typelib [23:33] they do that in the makefile.am [23:33] I'm wondering if that's their makefile breaking it [23:37] i would think it looks ok [23:37] at least that line [23:40] does gnome-shell have lots of depends i probably dont have installed yet? [23:40] not that much no [23:40] but it's over work hour so we can look at it tomorrow rather if you want [23:41] asac, deb http://ppa.launchpad.net/ubuntu-desktop/ppa/ubuntu karmic main [23:42] sudo apt-get install libclutter-0.9-dev gjs gjs-dev libmutter-dev [23:42] asac, basically [23:43] sudo apt-get install libclutter-0.9-dev gjs gjs-dev libmutter-dev libgirepository-dev gobject-introspection gobject-introspection-glib-2.0 gobject-introspection-repository [23:43] to be sure [23:43] but I'm not sure introspection is required [23:45] asac, the current wip package is on dget http://people.ubuntu.com/~seb128/gnome-shell_0.0.1~git20090701-0ubuntu0.1.dsc [23:46] seb128: are you building trunk? [23:46] or a branch? [23:46] asac, trunk [23:46] asac, git clone git://git.gnome.org/gnome-shell [23:47] for me the upstream stuff fails at http://paste.ubuntu.com/207789/ [23:47] http://paste.ubuntu.com/207789/ [23:47] err [23:47] Couldn't find include 'PangoCairo-1.0.gir' [23:47] asac, sudo apt-get install gobject-introspection-repository [23:47] asac, cf my second apt-get line before [23:47] indeed [23:49] http://paste.ubuntu.com/207791/ [23:50] next gir problem [23:50] cant find any Big gir here [23:50] asac, sudo apt-get install gobject-introspection [23:50] hmm. [23:50] i am sure i copied the line [23:50] thats installed already [23:51] ii gobject-introspection 0.6.3-0ubuntu1 [23:51] asac, the Big.gir is in the gnome-shell src [23:52] asac, you have /usr/bin/g-ir-scanner? [23:52] yes [23:52] its not in the src [23:52] anyway taking package now [23:52] clean your source and start again? [23:52] maybe it didn't find gir at configure time because it was not installed [23:52] and you did run make without running configure again [23:53] i reran configure [23:53] anyway [23:53] i am now building package ;) [23:53] ok [23:54] comment the export in rules otherwise it will work [23:54] fails the same way ;) [23:54] "work" [23:54] the Gir way? [23:54] http://paste.ubuntu.com/207795/ [23:54] thats with the export [23:55] bah [23:55] even without the export it fails the same way [23:55] so its _before_ the bug i want to look at [23:56] right [23:56] actually, this gir command is supposd to create this Big gir thing [23:58] asac, did you install mutter too? [23:59] seb128: i installed libmutter-dev [23:59] installing now [23:59] try installing mutter, in fact I built that one locally and dpkg -i *.deb [23:59] will mutter kill my metacity? [23:59] no [23:59] it's a different wm it will not run if you don't run it by hand