=== dduffey is now known as dduffey_afk [05:10] Good morning [05:11] pitti: word up [05:17] hey desrt, how are you? [05:24] laughing. [05:24] :) [05:24] just had a hilarious conversation with my girlfriend [05:24] she's doing an assignment using MPI [05:25] there's a function MPI_Send() that takes, as its first argument, an array of things to send [05:25] the type of the array is specified in a later argument, like MPI_INT or MPI_DOUBLE, etc. [05:25] anyway.. the C type of the first argument is void* [05:25] so it can take any type of array [05:26] but it's sending the array.... so why not const void * [05:26] anyway... [05:26] she had a function that had a 'const double *array' and wanted to send it using MPI_Send [05:26] and she was having a weird crash that she couldn't figure out... so i took a look [05:26] and i see MPI_Send (&array, .... [05:27] so i look at her and say "did it happen that you got a warning from the compiler about something to do with pointers and you thought to yourself 'hmmm.. sometimes these go away when i add a &, so let's try that...'?" [05:27] and she says "...HOW DID YOU KNOW?!" [05:28] (much laughing) [05:29] *chuckle* [05:29] remembers me when I spent about two hours explaining the concept of pointers in C [05:30] to my wife [05:30] it's funny because i remember the days when such a system used to work for me [05:30] for someone who previously only worked a bit with PHP, Java, and Python, this was quite a challenge indeed [05:30] "hmmm... compiler warning... i wonder if i should add more '*', or '&' instead?" [05:30] good to know I am in such good company :-) [05:31] desrt: yeah, it's like dealing with a nagging cat or so -- you try this and that randomly until it's satisfied :) [05:39] pitti: is there an equivalent to apt-cache show for source packages? [05:39] desrt: yes; apt-cache showsrc [05:39] lovely. thanks. [06:01] Morning === jono is now known as Guest29470 [06:38] robert_ancell: hey Robert, how are you? [06:38] robert_ancell: I'll move your remaining three a2 WIs to b1, ok? [06:41] * desrt discovers pbuilder --execute [06:50] pitti: Would you be up for sponsoring colord to Debian? [06:51] RAOF: sure! toss me the .dsc? [06:51] RAOF: oh, or from git, I suppose [06:51] Yeah, git would be nice :) [06:52] You'll need to uscan; pristine-tar is having a hissy fit about .orig.tar.xz again, it seems. [06:53] * RAOF thinks we should probably merge btrfs-tools from Debian, too; there's probably *something* interesting in the year and a half's development since our snapshot :) [06:54] good morning [06:57] bonjour didrocks [06:57] bonjour. [06:57] guten morgen pitti, how are you? [06:57] * RAOF scurries off. [06:57] hey RAOF :) [06:58] RAOF: weird issue i'm seeing lately, btw [06:58] RAOF: occasionally i'll be using my mouse scrollwheel in an app [06:58] then i'll go away and pay attention to another app for a bit [06:59] then go and try to scroll on the first app [06:59] and the result is that it seems that a whole lot of 'scroll up' events are sent at once [06:59] which either takes me to the top of a webpage or makes my volume very very loud [06:59] i've seen it happen at least with firefox and gtk3 apps [06:59] perhaps it's a weird symptom of the input transplant? [07:01] anyway... i'm having difficulty reproducing it, but it's happen maybe half a dozen times now [07:03] * desrt does bed [07:10] pitti, sorry, missed you there. That's fine, gtg, cya later [07:10] cu robert_ancell [07:17] RAOF: colord uploaded === warp11 is now known as warp10 [08:14] mvo: guten Morgen [08:19] hey, guten morgen pitti [08:38] good morning everyone [08:40] hey chrisccoulson [08:40] hi pitti, how are you? [08:41] quite fine, thanks! I'm happy to have bent aptdaemon and packagekit to my will :) [08:41] hey chrisccoulson [08:41] bonjour didrocks [08:41] 07:57:22 pitti | bonjour didrocks [08:41] spatial/temporal issue? :) [08:41] hey again pitti! ;) [08:42] didrocks: cerebral issue [08:42] didrocks: I'm just so happy to see you! [08:42] pitti: let's blame that on aptdaemon and packagekit! :) [08:42] pitti: ahah, I see that :-) [08:42] or on my DSL reconnect 10 mins ago [08:42] * didrocks hugs pitti [08:42] not sure whether my messages to mvo and GunnarHj came through [08:42] I don't see them [08:42] (irclogs.u.c. is behind) [08:43] didrocks: thanks, replaying [08:43] mvo: I recently landed plugin support in packagekit and aptdaemon, and would like to update aptdaemon to current trunk after alpha-2 [08:43] pitti: hi, I just saw a "hello" [08:43] pitti: +1 [08:43] mvo: this requires porting the s-c-aptdaemon-plugins to GI [08:43] mvo: I sent a MP, but do you have some general reservations about this? [08:43] pitti: I merged this this morning, many thanks! [08:43] GunnarHj: good morning! thanks for the fixes; reviewing now [08:43] GunnarHj: (and argh @ those UDD branches, I get merge conflicts) [08:44] mvo: oh, that was quick, thanks! [08:44] mvo: so we can upload that and aptdaemon after a2 freeze? [08:44] this works nicely now: pkcon what-provides "locale(de)" [08:45] so control-center or even our language-selector could use that upstream API now [08:45] yes [08:45] !!! [08:45] language-selector-common now just ships a five-line PK plugin which works in PK and aptdaemon-pkcompat [08:46] it's quite nice now [08:46] Good morning everyone [08:47] we could also extend this with other apps, i. e. app-install-data-ubuntu could ship a what-provides(application) plugin, or generalize our gstreamer handling [08:47] and I can drop most of jockey's smarts with that, too [08:47] morning everyone, [08:47] hey sil2100, hey agateau [08:47] got a strange error upgrading glib yesterday evening [08:47] http://paste2.org/p/1891666 [08:47] does it ring a bell to anyone? [08:47] yep [08:48] welcome to multi-arch [08:48] pitti: oh, that sounds interessting [08:48] (I bet) [08:48] pitti: I thought I entered multi-arch world last cycle? [08:48] agateau: did you dpkg -i a local build? [08:48] hey sil2100, salut agateau [08:48] agateau: it usually happens if the i386 and amd64 versions are not upgraded at the same time [08:49] pitti: I don't remember building glib myself, checking nevertheless [08:49] agateau: I also updated my amd64 machine from ubuntu-desktop PPA, that worked fine here [08:50] oh, I recently removed all :i386 packages to test the new glib, so nevermind [08:50] * sil2100 finished building and installing all those test packages [08:50] pitti: I don't have any i386 version of glib installed [08:50] pitti: could it be caused by i386 versions of other packages? [08:50] no, I don't think so [08:50] But I noticed that the compiz package you provided has a dlopen() of libplug1.so to /media/d1dddb1a-729f-40ef-9725-dc2a9ad56031/smspillaz/Source/Compiz/dev/dev/merges/compiz/core/compiz-core/build/src/libplug1.so [08:50] ;) [08:50] agateau: dpkg -S /usr/share/doc/libglib2.0-0/ChangeLog.pre-2-2.gz [08:51] agateau: that should say libglib2.0-0 [08:51] hey [08:51] bonjour seb128 [08:51] seb128: hi [08:51] pitti: mmm, that's embarassing [08:51] hey pitti, how are you? [08:51] pitti: libglib2.0-0:i386 [08:51] agateau: ah, so perhaps you updated at a time when the PPA didn't have the new i386 built yet? [08:51] seb128: I'm great, thanks! [08:51] pitti, agateau: hiting that gzip multiarch bug again? [08:51] we went bowling last night, quite some fun [08:51] seb128: yep [08:52] pitti, oh, nice ;-) [08:52] seb128: debugging ATM, I hope not [08:52] pitti, ricotz mentioning he was hitting it often with his ppa builds, he gave me a diff yesterday to exclude ChangeLog.pre-2-2.gz from the zippage [08:52] seb128: and I'm just boring everyone with the packagekit news :) [08:52] salut seb128 [08:52] seb128: oh, so it's that again? argh [08:52] lut didrocks [08:52] pitti, seems so [08:53] pitti: oh, just noticedd the libglib2.0-0:i386 as "iU" status in dpkg -l, [08:53] pitti: should i dpkg -r it? [08:54] agateau: it'll probably require you to uninstall other :i386 stuff [08:54] agateau: if you don't need it, yuou can sudo apt-get purge libglib2.0-0:i386 [08:54] pitti: ok, trying this [08:54] agateau: if not, just delete that changelog file from your system and dist-upgrade again [08:55] agateau: s/if not/if you need it/ [08:56] gaaaaaah, https://launchpad.net/~mozillateam/+archive/firefox-next/+build/3179551 !! [08:56] hi seb128 [08:56] how are you? [08:56] pitti: mmm, "apt-get purge..." => says to run apt-get -f install, "apt-get -f install" => complains about the .gz [08:56] hey chrisccoulson, I'm great thanks, how are you? [08:56] agateau, rm the .gz [08:56] pitti: going to rm the changelog file [08:57] agateau: bug 889303 FYI [08:57] Launchpad bug 889303 in gzip "gzip -9n sometimes generates a different output file on 64 bit (dup-of: 871083)" [Undecided,Confirmed] https://launchpad.net/bugs/889303 [08:57] Launchpad bug 871083 in gzip "gzip -9n sometimes generates a different output file on different architectures" [High,Triaged] https://launchpad.net/bugs/871083 [08:57] seb128, yeah, not too bad thanks. a bit tired though [08:57] pitti: ok, seems to fix it, thanks! [08:58] pitti: did I do something wrong to end up in this situation? [08:59] agateau: no, you didn't; it's the bug above [08:59] chrisccoulson, did you hack until early again? ;-) [08:59] pitti: ok [08:59] seb128, not really, but i'm stuck in a pattern of going to sleep at 4am ;) [09:01] chrisccoulson, that's enough to be tired I guess :p [09:04] and i'm nearly out of coffee! [09:07] slomo: hey, looks like gst-plugins-bad0.10 doesn't build a -dev package, needed for gstreamer-vaapi (GST_BASEVIDEO). would you accept a patch adding that? [09:11] tjaalton: no [09:11] tjaalton: there's no -dev package on purpose [09:11] tjaalton: the API of the -bad libraries is still unstable and possibly changes every release [09:11] sigh [09:12] so no way to build -vaapi then? [09:14] how often does -bad release anyway, the previous one was in may '11 [09:16] tjaalton: there's a pre-release from december and a new one will be there this week, and hopefully a final release (probably the last in the 0.10 series) end of next week [09:16] tjaalton: and no idea about gstreamer-vaapi, sorry [09:17] slomo: ok, gstreamer-vaapi is needed for 12.04 [09:17] tjaalton: "needed"? [09:17] tjaalton: perhaps you could copy some header files from -bad for the time being? [09:18] slomo: by OEM's [09:18] pitti: yeah, maybe.. [09:18] tjaalton: and then add a strict dependency, to force updating when -bad changes? [09:18] tjaalton: it's really not the brightest idea to depend on the gst-plugins-bad libraries ;) [09:18] but pitti's idea sounds good, even if it's still ugly [09:18] ok, I'll work around it [09:18] well, FSVO "good" [09:19] cowboy packaging FTW [09:19] :) [09:19] but if the next release is the last of 0.10, isn't it possible to add the -dev package for it? 0.11/1.0 will break the api anyway? [09:19] tjaalton: i'll get in contact with gstreamer-vaapi upstream and we'll try to find a solution [09:20] slomo: cool, thanks! [09:30] tjaalton: ok, if you give me a patch to gst-plugins-bad to add a -dev package with a shlibs file that will always generate a =${binary:Version} dependency on the package i'll upload that to debian [09:30] tjaalton: i guess that's the best solution for now, copying the headers is more likely to cause breakage [09:30] slomo: cool, will do [09:45] didrocks: is there a way to force debuild to run autoreconf ? [09:45] smspillaz: yes, there is a rather elegant way [09:45] smspillaz: for cdbs packages, include /usr/share/cdbs/1/rules/autoreconf.mk [09:45] smspillaz: for which package? (depends on which debhelper version) [09:45] smspillaz: for dh7 packages, dh --with autoreconf [09:46] didrocks: metacity [09:46] smspillaz: and add a dh-autoreconf build dep [09:46] it doesn't run autoreconf :( [09:46] smspillaz: you need to add a build-dep on dh-autoreconf [09:46] ok, will do [09:46] metacity is cdbs [09:49] pitti: make: *** No rule to make target `/usr/share/cdbs/1/rules/autoreconf.mk'. Stop. [09:49] dpkg-buildpackage: error: fakeroot debian/rules clean gave error exit status 2 [09:49] debuild: fatal error at line 1350: [09:49] smspillaz: as didrocks and I said, you need to add a dh-autoreconf build dep [09:49] that package ships it [09:50] ah fun [09:52] pitti: Hello, Martin! [09:52] hey GunnarHj [09:52] pitti: Merged? Yeay!! :) [09:52] GunnarHj: heh, it took a while :) [09:53] mvo, pitti: do you have any opinion on what to do about bug #923171 [09:53] Launchpad bug 923171 in update-manager "software-center crashed with AttributeError in popup(): 'Menu' object has no attribute 'popup_for_device'" [High,Confirmed] https://launchpad.net/bugs/923171 [09:53] GunnarHj: I'm currently debugging the migration in postinst, it also writes everythign to /etc/environment [09:53] seb128: will look in a bit; wit gtk 3 popup_for_device() is supposed to work [09:53] mvo, pitti: it's basically the gtk currently in the ubuntu-desktop ppa which has an api break in annotations [09:53] pitti: Yep. l-s should be much easier - in effect no important changes. [09:54] pitti, well, upstream renamed it on purpose, it's not really a bug, but it's still an api break... desrt started a discussion on the gtk devel list about binding stability [09:54] has anyone experience with the Geoclue dbus interface? when I use the ubuntu provider directly it works fine, but when I use org.freedesktop.Geoclue.Master it tells me it has no address provider [09:54] seb128: what are the options? [09:54] pitti, mvo: meanwhile not sure if we should revert or do something else or fix update-manager and software-center [09:54] mvo, ^ those 3 I guess [09:54] pitti: Re: postinst - yes, shouln't it? That's what l-s has done all the time. [09:55] 1- revert the rename, but that makes us api incompatible with upstream, I don't like it [09:55] 2- somewhat add the old name as an alias or something (not sure how it can be done technically) [09:55] 3- fix the rdepends [09:55] ah, I see [09:55] fix -> update [09:55] seb128: I'm fine changing the code in s-c/u-m, its just yet another bad experience for me as a develop (and for others) [09:55] I think (2) is the best option for upstream [09:55] right [09:56] but *shrug* [09:56] smspillaz: what does compiz use libplug.so for? [09:56] seb128: with the rename-to: I'm not sure whether we can write an override (your option 2) [09:56] mvo, if you can get 3) meanwhile, I plan to upload gtk after the a2 freeze [09:56] mvo, so I would like to not break stuff and you to be ready ;-) [09:56] pitti, yeah, me neither :-( [09:56] seb128: what needs to be used instead? [09:56] seb128: I'll examine 2 after I'm done with accountsservice [09:56] pitti, danke [09:56] mvo: just popup(), as in the old days [09:56] ok [09:57] what pitti said [09:57] but of course if I do that now it will break with current gtk [09:57] right? [09:57] yes [09:57] yes [09:57] (Geoclue anyone ;) ? [09:57] *sigh* [09:57] :-( [09:57] ok, thanks [09:57] mvo: I'll investigate re-inroducing _for_device() [09:57] I guess you can do a "get gtk version and conditional code paths" [09:57] thanks pitti [09:58] do we have any idea for how long this was actually available? [09:58] smspillaz: since main.cpp in compiz does a dlopen of libplug.so with some absolute path from your system, hm [09:58] mvo, but I agree it sucks hard, well at least this time they have an upstream discussion on how to solve this problem [09:58] i mean, what timeframe are we talking about? [09:58] mvo, for what? [09:58] sil2100: um. whooooooooops [09:58] popup() vs popup_for_device() [09:58] mvo, the rename is in the gtk which is in the ppa that will go to precice when the a2 freeze lifts [09:58] I think that was the tarball I was using to debug the linking bug [09:58] err [09:58] I assume that this is used by more than just u-m/s-c [09:58] I'll get you a new one [09:58] mvo, so I hope today [09:59] I mean, when was popup_for_device introduced, how long ago? [09:59] mvo, well it's only an annotation rename, it doesn't affect C or vala [09:59] can we hold this back a bit? [09:59] smspillaz: hoho, so should I just remove the dlopen() thingy ;)? Or is there a bit more I'll have to remove? [09:59] (this is what happens when I rush around) [09:59] smspillaz: thanks ;) [09:59] sil2100: uhhh, I'll get you a new one, hang on [09:59] seb128: it's also used in computer-janitor and other places [09:59] pitti, sure, anyway I just wanted to start the discussion early so maybe we can resolve it and not delay the gtk upload too much [10:00] I still don't understand http://bugzilla-attachments.gnome.org/attachment.cgi?id=206147 [10:00] popup() and popup_for_device() have two distinct APIs [10:00] pitti, but deal with your accountsservice changes and ping me back later when you have time for it [10:00] why can they rename one to be the other? [10:00] pitti, well see https://bugzilla.gnome.org/show_bug.cgi?id=657385 [10:00] Gnome bug 657385 in menu "[introspection] gtk_menu_popup() not introspected" [Normal,Resolved: fixed] [10:01] they expose only one to bindings on purpose it seems [10:01] pitti, but anyway, finish what you are working on [10:01] we can discuss gtk after lunch or later today [10:01] yes, but popup_for_device() requires a device argument [10:02] heh :) does that mean that if you lookup how to use it using the C docs you get the wrong info? not quite the best approach [10:02] I mean, there are no binding specific docs and now when you look at the C docs they tell you something wrong? [10:03] yes, that's my concern [10:03] mvo, pitti: ok, what about reverting the annotation change for this upload? so we don't block gtk and don't break anything until we sort things [10:03] sil2100: rebuilding hang on [10:03] mvo, seb128: also, it's not a concern for python [10:03] mvo: you can use .popup() right now [10:04] there's an override which defines a python .popup() bycalling popup_for_device(None, ...) [10:04] seb128: fine for me; it seems actively wrong to me [10:05] pitti, mvo: ok, I will revert for that upload so we can sort it quietly without rush [10:05] thanks [10:06] seb128: is there a good place to ask about geoclue? [10:06] mvo, not sure, you can try pinging bastien on #gnome-hackers if you have the skin for it :p [10:07] smspillaz: ok ;) [10:07] mvo, otherwise wait for kenvandine to join, he's probably the one knowing the best that stack around, he worked with it before [10:07] seb128, mvo: I followed up to the upstream bug [10:07] seb128: I will wait for ken [10:07] seb128: I don't have enough guts for flames this morning [10:08] mvo, ;-) [10:08] pitti, danke === zyga_ is now known as zyga-xchat [10:25] GunnarHj: sorry, DSL reconnect; if you said something in the last few mins, I didn't get it [10:25] hey seb128 quick ? for ya. I am testing a new install on a new machine and I am trying install dropbox, but I'm getting an error. [10:26] GunnarHj: so, got everything queued up in bzr, now waiting for the freeze to lift [10:26] jasoncwarner_, hey [10:26] seb128: nautilus-dropbox depends on libnautilus-extension1 (>= 1:2.22.2); however: [10:26] Package libnautilus-extension1 is not installed. [10:26] seb128: hey :) [10:26] jasoncwarner_: sounds like a dropbox version for lucid or so [10:26] when I try to install that extension, I can't b/c it is replaced by libnautilus-extension1a, which also doesn't install? [10:26] jasoncwarner_, seems that version of nautilus-dropbox needs to be rebuilt for precise [10:26] jasoncwarner_: isn't that spelled "ubuntu one"? :-) [10:27] pitti: :) yeah, it is [10:27] pitti: I still have some old docs on there, unfortunately it seems [10:27] I thought the dropbox guys even asked us to remove dropbox from the archives [10:27] jasoncwarner_, the current version has a "a" at the end indeed, nautilus-dropbox needs to be rebuilt to use that one [10:27] pitti: I tried downloading .deb from dropbox.com [10:27] seb128: ok...I'll ask them to do that ;) [10:28] jasoncwarner_, ok [10:30] hah: http://status.ubuntu.com/ubuntu-precise/canonical-desktop-team-precise-alpha-2.html [10:30] (ok, cheating: moved some WIs) [10:30] but we are clean wrt. alpha-2 now [10:31] :) [10:32] pitti: I think I'll have to drop my remaining oneconf WI (with the installer) as there is no work on ubiquity this cycle and with the latest additional requirements from design for g-c-c [10:33] didrocks: ah, too bad :( I know it's a project close to your heart :( [10:33] pitti: yeah, but it's always postponed cause ENOTENOUGHTIME or EDEPENDSONOTHERTEAM :/ [10:34] (and the isd team dropped their oneconf WI) [10:34] well, at least, we will still have the software-center integration :) [10:35] (at least, it gained a full testsuite this cycle) === Guest50763 is now known as p === p is now known as Guest62719 [10:55] fantastic. whilst investigating another bug, i've discovered that the current firefox nightly doesn't build on armel :-( [11:16] Er.. How do I choose unity-2d in Precise? I don't see a session chooser in the Precise lightdm. [11:17] I like the love from the testers [11:17] "this test is silly, F is not the accelerator in all languages for File" [11:17] (2 people) [11:17] rather than telling that, please help updating it! :) [11:18] I see that unity-2d is installed by default, but how do I choose it? [11:18] soren, you click on the small gear icon on the greeter [11:18] next to your name [11:18] It [11:18] It's not there. [11:18] I was in Oneiric.. [11:18] soren, update your unity-greeter to the currnet version and restart? [11:18] * soren goes and checks again [11:19] soren, I fixed that tuesday [11:19] soren, well I backported the fix from mterry to be exact [11:19] Note: This is not an upgrade (it's a new laptop), so I may be missing some packages or whatnot. [11:19] seb128: Oh. Yeah, I'm a couple of days behind on updates (hotel wifi and whatnot). I'll update. Thanks. [11:20] I just assumed the functionality moved somewhere else [11:20] soren, unity-greeter is tiny, otherwise log in and log out or switch user, I think it was off only on first greeter start [11:37] seb128, pitti, etc... [11:37] A2 today, what's the word on the street? [11:37] didrocks, mot dans la rue? [11:38] rickspencer3, no issue that I know about [11:38] but I didn't follow things closely out of desktop land [11:38] rickspencer3: unity 5.2 will maybe be slightly delayed (boot time regression). I'm continuing collecting the user results, taking time :) [11:39] hmmm [11:39] isn't A2 image already rolled? [11:40] rickspencer3, yes, unity 5.2 will be uploaded after a2 [11:40] rickspencer3, i.e orthogonal topic [11:40] ackity [11:40] right, thanks seb128 === Guest62719 is now known as mandel [12:00] jasoncwarner_, you're using the aurora builds from https://launchpad.net/~ubuntu-mozilla-daily/+archive/firefox-aurora/+packages aren't you? [12:01] chrisccoulson: I was on my other machine..but I had to send ti back to get repaired...on this new one I'm just using stock firefox [12:01] chrisccoulson: firefox 10 [12:01] jasoncwarner_, oh, never mind then :) [12:01] chrisccoulson: you want me to test something? [12:01] i was going to ask if you wanted to smoketest the new builds, before i turn publishing back on again [12:09] soooooo, when can we start uploading again? :) [12:20] rickspencer3: haven't really followed A2 this time, I'm afraid [12:21] it's been fairly quiet today, as most of the action happens during US day [12:25] can we upload yet?! [12:25] can we upload yet?! [12:25] can we upload yet?! [12:25] ;-) [12:26] smspillaz: how's the rebuilding proceeding ;)? [12:26] I mean, re-packaging [12:26] oh right [12:26] whoops [12:26] I'll upload the new tarball now [12:26] sorry got caught up in a bunch of stuff [12:27] sil2100: same place as it was last time [12:27] (keep in mind I haven't had a chance to thoroughly test this one) [12:27] at least all the unit tests pass though [12:28] seb128, hi [12:28] ricotz, hey, how are you? [12:28] smspillaz: thanks - sorry for the trouble! [12:28] sil2100: no, thank /you/ :) [12:28] seb128, great, you too? [12:28] ricotz, I'm fine thanks [12:28] and /sorry/ from myself from the delay and the accidental oops [12:29] seb128, did you added the glib packaging changes? [12:29] ricotz, no, I plan to do that for the upload to precise once the a2 freeze lifts which is not done yet [12:29] especially libelf-dev [12:29] alright [12:30] ricotz, I didn't plan to do another ppa upload only for that [12:30] by the time the ppa build the archive will probably be unfrozen :p [12:30] i see ;) [12:31]