=== Zachary_DuBois is now known as Zachary_DuBois|A === happyaro1 is now known as happyaron === cmiller_ is now known as CardinalFang === CardinalFang is now known as qengho [09:04] guten morgen! [09:04] morning Laney [09:05] good morning desktopers! [09:05] Laney, hey, how are you? had a good w.e? [09:06] hallo, yeah it was good - saturday was sunny so I roamed around outside doing various things but sunday it rained all day [09:06] = playing games and making sloe gin [09:07] you? [09:08] saturday was raining, sunday sunny for us [09:08] I did some errands/shopping on saturday, some "house-cleaning as well, and played tennis in between [09:09] yesterday mostly enjoyed the weather and spent some time with my cousin who is here for the holidays [09:09] today it's raining again :/ [09:10] but they forecast 16°C and sunny for next w.e [09:10] crazy weather! [09:10] seems we get the opposite weather to you [09:10] blue sky outside now! [09:10] bah, give it back! [09:11] ;-) [09:11] oh, seems like the "download fixes" landing from friday was buggy :/ (it made clicks unhappy) [09:11] yeah I saw [09:11] apparently nobody did the full test plan [09:12] I guess everyone checked the system update bit [09:12] whoops [09:12] I didn't even realize that was doing clicks [09:12] yeah === MacSlow is now known as MacSlow|lunch [11:40] mlankhorst: hey, I was checking https://bugs.launchpad.net/ubuntu/+source/unity/+bug/1055166 [11:40] Launchpad bug 1055166 in unity (Ubuntu) "compiz crashed with SIGSEGV in memmove() from drisw_update_tex_buffer() from dri_set_tex_buffer2() from operator() from compiz::opengl::bindTexImageGLX() from ... from unity::UnityWindow::DrawWindowDecoration" [High,Triaged] [11:40] mlankhorst: and it seems that the problem happens when a texture is 0x0... [11:41] mlankhorst: in that case I'm getting at some point on the bt: http://pastebin.ubuntu.com/7026994/ [11:41] where w/h seems to have random values [11:42] mlankhorst: fyi the problem is solved by not creating textures of null size, but it would still be something that should not crash btw [11:46] Trevinho: but if I add a flush I don't get a crash [11:46] mlankhorst: I don't know how that might fix it... mhmh [11:46] mlankhorst: flushing after building the pixmap? [11:47] flushing before the failing XGetGeometry call [11:48] mlankhorst: ah, yeah so basically between the pixmap creation and the binding of the pixmap to texture... [11:48] in the same way, btw, if I don't create a texture if the size is 0x0, it works... mhmh [11:54] mlankhorst: at unity level flushing after creating the pixmap doesn't seem enough tohg [11:54] tough* [12:03] Trevinho: yeah I did a flush at the crashing call, so I have no idea what it is, but I'm not really sure that it's a bug in libdri [12:03] is any error suppressing libx11 handler installed btw? [12:03] might be good to find out the first real error [12:04] mlankhorst: mhmh, herrors should be catched by gdk I think here... let me check [12:06] mlankhorst: oh, yeah... we have in compiz let me get the info [12:09] mlankhorst: I'm only getting a couple of BadWindow error on X_GetProperty calls [12:09] ah, no wait I've the fixes installed :) [12:09] does anything happen around the time of crash? [12:12] mlankhorst: no, even with crashing unity, still no x errors [12:13] the XGetGeometry call fails, and afaict the only failure allowed is BadDrawable [12:13] surely that error ends up *somewhere* ? [12:14] mlankhorst: mh, we have an error handler in compiz, that does nothing by default (unless enabling debug), then nothing else [12:14] I think there's not a gdk error handler set, but it shouldn't interfere [12:15] :\ [12:16] well XGetGeometry is failing before the real crash, that should at least generate *some* kind of error.. :/ === vrruiz_ is now known as rvr === alan_g is now known as alan_g|lunch [12:44] bregma, hey, what's the plan for the unity hi-dpi fixes landing? ;-) === tkamppeter__ is now known as tkamppeter [12:52] Trevinho, i noticed that with latest compiz window placement doesn't seem to take the titlebar into account, known? === MacSlow|lunch is now known as MacSlow [12:54] mhr3_: known and fixed [12:54] Trevinho, in distro? [12:54] but not in distro yet [12:54] i see [12:55] Trevinho, what about my shadow issue? :) [12:55] mhr3_: it's on my list... but I've to finish other things before... :) [12:56] Trevinho, is there a workaround possible? [12:56] mhr3_: for synapse you mean? [12:56] mhr3_: set it as argb window [12:56] how would i do that? [12:58] Trevinho, is that a window hint, ccsm setting or...? [12:58] mhr3_: let me check the higher level call :) [13:00] mhr3_: ah, no the argb thing won't work... as it already is :) [13:00] so.... mhmh, try to mark it as a dock type window for now [13:00] is the easiest workardound [13:06] Trevinho, hm, looks like it was using SPLASHSCREEN as a special case for compiz [13:06] breaking existing workarounds! evil Trevinho! [13:06] ahah :D [13:14] mhr3_: anyway to get the proper fix I think you should instead use a cairo mask to shape the window to match the style,,, so you'd get events only in that area and unity won't draw any shadow :) [13:17] Trevinho, hm, seems like we're doing that for kwin already [13:18] Trevinho, i guess there were other issues so we had to use non-NORMAL window type with compiz [13:21] Trevinho, oh maybe it was just to get rid of the stupid appear animations :) [13:23] Laney: good morning [13:24] Laney: looked into that one remaining tests failure over the weekend. it's a hardware bug :) [13:29] charles: Hi, can you please re-review https://code.launchpad.net/~mitya57/indicator-applet/unparent-label/+merge/208174 ? === alan_g|lunch is now known as alan_g [13:32] Trevinho: Oh, are you the reason gnome-do has recently picked up some shadows on its hidden results window? :) === psivaa is now known as psivaa-lunch [14:07] Hi happyaron, around? [14:11] Laney: but seriously... some more info about which machine that test failed on would be helpful because it looks like it's impossible for it to fail unless the machine has some very strange floating point implementation (non-standard rounding rules, non-deterministic behaviour, etc.) [14:11] if it's a softfloat machine then maybe we found some problems there... [14:16] tseliot, Sarvatt who should be bugged about bug 1285205 / bug 1152656? Not sure this is known and tracked as something else or not on the radar yet. Looks to be present after todays updates, still for me. [14:16] Launchpad bug 1152656 in compiz (Ubuntu) "compiz crashed with SIGSEGV" [Medium,Confirmed] https://launchpad.net/bugs/1152656 [14:17] smb, Trevinho has a fix for it [14:17] smb: not me, no proprietary drivers there ;) [14:18] smb, https://code.launchpad.net/~3v1n0/compiz/ignore-invalid-sized-textures/+merge/209061 [14:18] seb128, Ah ok. Thanks [14:19] smb, yw [14:25] desrt: they were either jenkins slaves (VMs?) or archive builders [14:25] sadly the old logs have gone away [14:26] tsk. [14:26] I could turn the tests back on ... [14:26] Laney: so i'm kinda convinced the tests aren't buggy [14:26] like the one failure of assert (a == b) where it showed that the two were equal... [14:27] this test consists of performing the same calculation twice and making sure it comes out the same both times [14:27] ...which apparently it doesn't [14:27] pretty dodgy... [14:28] the other failure is .. possible to imagine, at least. it consists of someone rounding improperly. [14:29] seb128, Hm, ok. Looks to me like something there causes ci to fail (or something else) [14:30] smb, right [14:30] bregma, btw, whenever you do an unity landing, could you try to include https://code.launchpad.net/~3v1n0/compiz/ignore-invalid-sized-textures/+merge/209061 ? it's currently creating issues for qa tests (segfault in vm/using software rendering) [14:31] smb: seb128: We are working on straightening out the CI failure and then once that's done, we'll review and get it on the train. [14:31] ChrisTownsend, thanks [14:31] \o/ seeing Trevinho's hi-dpi branch on the train! [14:31] seb128: np [14:32] desrt: I'll turn them back on in the next release [14:32] ChrisTownsend, Ok, thanks. [14:32] expecting a tarball today :-) [14:33] Laney: yup. will be releasing it. [14:33] have a few fixups to push first, though. [14:33] Trevinho: :) === psivaa-lunch is now known as psivaa [14:47] seb128, I see the “Prompt when not connected” feature isn’t implemented yet. What package would it belong to? [14:48] mpt, indicator-network and ubuntu-system-settings I would say [14:49] tedg, Wellark: ^ right? [14:50] seb128, Yes, indicator-network mostly. The code exists, it just needs to be integrated. [14:50] Right now it's in indicator-network-prompt, but it'll be moved into indicator-network. So the bug should go there. [14:51] tedg, thanks [14:57] Laney: how come gstreamer0.10 still on the cd? [14:58] is it just seeded or actually needed for something? === mjohnson15_2 is now known as mjohnson15 [14:59] check for rdepends on the image [14:59] of the library [15:04] Laney: found one, bluez-gstreamer. [15:04] all others appeared to be circular back to ubuntu-desktop. [15:05] xnox, Laney: ubuntuone stuff use python-qt4 which depends on libqtwebkit4 which depends on libgstreamer-plugins-base0.10-0 [15:05] ubuntu-sso-client-qt [15:05] ubuntuone-control-panel-qt [15:06] seb128: oh, i see. about that, do we have anybody to port that stuff to qt5? [15:06] xnox, not that I know... [15:06] to python3 as well would be nice [15:06] seb128: ack. [15:08] mpt, JohnLea: hey you two! I sent out an email about sharing the volume with the greeter last week. If you get a chance, I'd like feedback on what ya'll think about the use cases. I believe I'm going to go ahead and work on a patch to share it, but would appreciate guidance [15:09] * xnox tries to remember Mathieu Trudel-Lapierre real irc name..... [15:09] lol [15:09] cyphermox [15:09] yup [15:09] cyphermox: hey! any plans to switch to bluez 5.x ? =))) [15:09] xnox: nope [15:09] xnox: why? do you need something from it? [15:09] cyphermox: how about disabling gstreamer compile option? [15:09] sure. [15:10] IIRC we don't install that stuff usually, though [15:10] cyphermox: 5.x doesn't have gstreamer =) [15:10] I know [15:10] what uses bluez-gstreamer currently though? :) [15:10] $ seeded-in-ubuntu -b bluez-gstreamer [15:10] bluez-gstreamer is seeded in: [15:10] edubuntu: dvd [15:10] ubuntu-gnome: daily-live [15:10] ubuntu: daily-live [15:10] ubuntukylin: daily-live [15:11] cyphermox: for one reason or another it's seeded into all of above.... [15:11] right [15:11] xnox, cyphermox: what feature does that provide? [15:11] I'm not sure it's actually used by anything, but we'll have to look into it [15:11] xnox, what's the point to drop it we can't drop gst0.10 anyway due to u1? [15:11] some kind of gstreamer library ;) [15:12] seb128: i have a stronger bargaining chip =) [15:12] (with security team) =)))) [15:12] xnox, I doubt we are going to drop u1, that's a bargain you are going to loose [15:12] xnox: I'll look into it, since I'm messing in bluez anyway [15:13] seb128: or win development time to port it. [15:13] good luck [15:13] cyphermox: seeds just says "Bluetooth: recommend gnome-bluetooth, and bluez-gstreamer" [15:14] cyphermox: added in 2010 by pitti, "explicitly add bluez-gstr4eamer, desktop-common dropped it" === JanC_ is now known as JanC [15:29] mterry, sorry I was out a chunk of last week. I’ll have a look at it tomorrow morning. [15:30] mpt, no crazy rush. I think the technical work is pretty similar regardless [15:30] mpt, still having those unity-control-center strings suggestions on your todolist as well? ;-) [15:30] yessss [15:31] great, sorry for piling on all those requests you already have [15:32] attente: Hey, when you have a sec, let's discuss https://code.launchpad.net/~attente/compiz/1284532/+merge/208908 [15:33] ChrisTownsend, sure [15:34] attente: So I made a comment, but then thought maybe my comment is not so valid:) [15:34] ChrisTownsend, no, it's a concern, definitely [15:35] attente: Do you know if other downstreams (such as Gnome Flashback) might still use this plugin? [15:35] Trevinho, oh, that ^ change interests you [15:35] ChrisTownsend, no idea unfortunately... [15:36] i guess alternatively we can just change the defaults for those settings to null them out [15:36] but that's also a distro-patch [15:36] attente: Ok, well, maybe it's best to just distro patch disabling the plugin when using Unity. [15:37] attente: Leave the other code intact. [15:37] ChrisTownsend, i think there are two other settings there that are still used by unity [15:37] attente: Just in case somebody wants to use it. === greyback is now known as greyback|rebooty [15:37] "show main menu (alt+f1)" and "run dialog (alt+f2)" [15:37] attente: Oh, right, I mistyped. [15:38] attente: I meant disable the settings as you said. [15:38] ChrisTownsend, ok, thanks, sounds reasonable [15:38] attente: Cool. Let me know when your MP is ready and I'll review it. [15:40] Laney, I think you should put a landing for https://code.launchpad.net/~laney/ubuntu-system-settings/as-ringtone/+merge/200862 [15:40] but I'm scared [15:40] ;-) [15:43] haha [15:43] I will do, let me do this reset api fix and then do that [15:44] Laney, that v2 thing is the plugin api confuses me [15:44] what is needed for a plugin to declare it's on v2? [15:45] change the interface it implements [15:45] can we warn when a plugin is not on v2 to avoid confusion? [15:45] and the Q_SOMETHING lines [15:45] umm, don't think so [15:46] what change is needed to -online? can we just do a synced landing of both? [15:46] it feels like we are arguing about a theoritical issue since we have 1 external plugin known and we control it [15:46] I already proposed a branch for it [15:46] so we can jdi [15:46] but mardy still asked for this [15:46] and you backed him up, so I'm doing it now [15:47] well, I don't really understand what are the issue, it just seemed that the check was low cost and safer [15:47] but I didn't think it could create issues and create confusion [15:48] It's just that you can't reset if you aren't 2 [15:48] We can probably remember that [15:48] well, that's fine I think [15:48] honestly I don't see any issue either way [15:49] I just wanted to let people reset from javascript whenever [15:49] but it doesn't really matter that much [15:49] right [15:50] Laney, not sure why mardy insist on that, he seems like he's ok to drop those lines later [15:50] we can as well do a synced landing, pinged him about that on #ubuntu-devel [15:50] I don't know why he said that [15:50] Either it's a good idea in which case keep it or it isn't in which case don't do it at all [15:52] right [15:52] well, I would go for a synced landing [15:53] shrug, I've done it now [15:53] but I want to hear back from mardy before deciding [15:53] sorry about that, really I don't care much, they are all equal options :-) [15:53] :P [15:53] just pick one [15:53] actually the test broke [15:53] noooooooooooooooooo [15:53] :/ [15:54] oh it was this exact issue [15:54] haha [16:44] bregma, can I press "build" for the unity7 silo or do you hold on for a reason? doing a build now doesn't block you from redoing one later but it would give debs to those of us who want to test the update [16:45] seb128, oops, go ahead, I got buried by other stuff and forgot to do it myself [16:46] bregma, done [16:46] thanks [16:46] bregma, I commented on the unity8 session mp with a "you might to fix that" [16:47] seb128, yes, it's in my to-do queue [16:48] cool === robru is now known as robru-at-doctor [17:53] bregma, ChrisTownsend, Trevinho: the unity update from ppa-004 works nicely for me, good work ;-) (just giving feedback since it's up for testing) [18:08] * didrocks waves good evening and good night [18:53] cyphermox, hey, do you know if ofono provides what we need for https://bugs.launchpad.net/ubuntu/+source/ubuntu-system-settings/+bug/1287267 ? [18:53] Launchpad bug 1287267 in ubuntu-system-settings (Ubuntu) ""Data usage statistics" not shown" [Low,Triaged] [18:53] cyphermox, e.g stats on datas sent/received [18:59] seb128: Cool, thanks for the feedback:) [19:40] mdeslaur, gnote?! [19:44] seb128: yes! no? [19:44] am I the only gnote user? :) [19:44] mdeslaur, well, patches are welcome, but we don't intend to maintain patches for !default_install [19:44] I see [19:44] the only one that cares about menus I guess [19:44] ;-) [19:45] * seb128 uses tomboy, but can see the interest in not having mono installed [19:45] it's in universe, I assumed some community member would maintain it :) [19:45] yeah, if somebody comes with a patch, fine, we are just not going to write [19:45] +it [19:45] I took a quick look, but it's C++, so I got a nosebleed [19:46] seb128: oh, I wasn't expecting you to [19:46] I'm already happy we managed to get nautilus evince file-roller and some others [19:46] yes! [19:46] larsu is our hero there ;-) [19:46] * mdeslaur hugs larsu [19:46] * seb128 hugs larsu as well [19:48] * larsu feels hugged [19:50] ;-) === robru-at-doctor is now known as robru-sick === mjohnson15_2 is now known as mjohnson15 === m_conley_away is now known as m_conley