[00:56] chrisccoulson, hey, got a minute? [00:57] nigelb - possibly ;) [00:57] chrisccoulson, after the epiphany-browser update, the epiphany-extensions and epiphany-extensions-more dont work [00:58] they probably need to be sync'd from debian too [00:58] epiphany-extensions need to be, epiphany-extensions-more isn't updated in debian yet [01:00] chrisccoulson, so you'll by syncing it ? [01:01] nigelb - i can't sync packages, but you could request a sync [01:02] chrisccoulson, aren't we in final freeze ? [01:02] ugh [01:02] we will be in a minute ;) [01:03] ouch, so isn't it going to be twisted since I need everything all the way from FFE? [01:05] I didn't get my gnome-bluetooth bug fixed :( [01:06] Nafai, bug fixes go in just fine :) [01:07] oh good :) [01:07] new to the process here [01:08] only someone has to manually approve it :) [01:09] regardless, I don't yet have upload permissions so someone else would have to take care of it anyway :) [01:09] Nafai - whats the bug? [01:10] https://bugs.launchpad.net/ubuntu/+source/gnome-bluetooth/+bug/558841 [01:10] Launchpad bug 558841 in gnome-bluetooth "bluetooth "devices" menu item not working in bluetooth indicator" [High,In progress] [01:10] the devices aren't showing up on the indicator [01:10] I believe it is actually a bug in appindicator libraries [01:10] and I have narrowed it down to the release where it is exhibited [01:10] but it is confusing because there doesn't seem to be code changes that might cause that [01:17] rickspencer3, http://insidesocal.com/click/2010/04/im-running-the-ubuntu-1004-bet.html [01:20] * rickspencer3 looks [01:21] "I've heard/read that the ability to turn off kernel mode setting might either be changing or going away in future kernels. I have nothing confirming or denying this. I only know that the Xorg developers really don't care about what users of older Intel video chips have had to go through for the past couple of years. [01:21] " [01:21] rickspencer3: also, https://bugs.launchpad.net/ubuntu/+source/gnome-bluetooth/+bug/558841/comments/7 :) [01:21] ouch [01:21] Launchpad bug 558841 in gnome-bluetooth "bluetooth "devices" menu item not working in bluetooth indicator" [High,In progress] [01:22] bryceh, I assume that refers to upstream, not us ;) [01:23] rickspencer3, right [01:24] rickspencer3, it's a good signal that our decision to disable kms on 830 is going to be well received [01:24] mmm [01:24] nice [01:24] Nafai so you're a bit blocked? [01:24] have you debugged into the code and found where the UI states are set and such? [01:29] yeah [01:29] (to both) [01:30] and they definitely are set correctly on the bt side, but aren't caught over on the appindicator side [01:33] so the app indicator is not seeing the messages? [01:34] correct [01:35] but the messages are being sent? [01:35] what is different about the bluetooth messages? [01:35] well, the menu structure is being set, but the watcher over in libappindicator isn't picking up the changes [01:36] hmm [01:36] but is the applet ripping the dbus calls? [01:37] I'm not sure I follow [01:38] how does the gnome-bluetooth tell dbusmenu that there are devices present? [01:38] what might be different here is in this case after the menu is set on the app indicator in the bt indicator, actions are added for each of the devices (when you turn on bluetooth and it sees each of the devices). these GtkActions are mapped to menu items [01:38] libappindicator has some kind of proxy that listens to signals from the GtkMenu you have set and then makes appropriate calls to dbusmenu [01:39] in this case, libappindicator isn't getting whatever the appropriate signal might be and thus isn't making the dbusmenu calls [01:40] now, the other menu items that show up when you turn on bluetooth are already on the menu, but are just not visible [01:40] so libappindicator is picking up the visibility toggle signal [01:40] for those items [01:42] Nafai, can you explain to me briefly how the indicator discovers available devices? [01:43] sure, let me double check the code real quick to check my understanding [01:47] ok. there is bluetooth client code that underneath uses bluez (I imagine) to listen to the actual bluetooth adapter. it is hooked up to a GtkTree model. When a row is added to this model via the client (i.e., the adapter sees another device), a callback is called that will update device list by iterating through the tree model, adding submenus and menu items based on the capabilities and type of the particular device [01:48] it adds these to a GtkUiManager, btw [01:49] how does the client hear from bluez? [01:49] for example, that a device is added? [01:52] is there something wrong with the i386 builders.. or is there just THAT many packages waiting to be built? [01:53] bcurtiswx, 869 jobs in the queue [01:53] that I'm not sure, that's in code separate from the applet, looks. I've taken for granted that it works, which my debug prints show that the devices are seen [01:53] chrisccoulson: Gotta get things in before the freeze! :) [01:53] and one of them is building openjdk-6, which is pretty big [01:54] 27 hours [01:54] ouch [01:54] chrisccoulson: interesting.. so tomorrow i should expect a hail mary worth of packages then [01:54] probably ;) [01:54] wwe [01:54] er, wee [01:55] Nafai. lol [01:55] Nafai there will still be time to get the fix in [01:56] so the client sees the devices, have you confirmed that the TreeModel is updating properly? [01:56] good [01:56] anyone here follow the NHL? [01:57] yeah, or the update_device_list wouldn't be called. [01:57] what object has update_device_list? [01:57] I honestly think (and unfortunately nothing in the difference between the last working and the broken version helps me here) that there is something not being seen about the action that gets added for each device [01:57] Nafai, yeah [01:57] I'm trying to trace teh data through [01:58] so can you confirm the treemodel is getting updated *properly* [01:58] and that the right data is getting sent? [01:58] oops dinner ready [01:58] bbiab [01:58] yeah, I'll do that next while you are eating :) [02:01] what is a "private source" ? [02:12] back [02:13] this would be easier to debug in Python :) Easier to display data structures, etc :) [02:14] for surez [02:15] let me add one more debug thing to make sure the menu is getting created correctly. I've only got a debug print where it gets added, but I'm not actually showing it [02:31] Nafai - you find python easier to debug? [02:32] in some ways, definitely [02:32] because of introspection [02:32] i never know what tools to use to debug python properly [02:33] print :) [02:33] i suppose i just like C far too much ;) [02:33] gdb FTW :) [02:33] I've never really used a python debugger [02:33] well, the problem I'm having is I'm trying to examine a GtkMenu structure [02:33] no debugger will help much with that [02:33] in python, I'd just print it and the str representations would normally give me all I need [02:33] the data structure will be pushed way down into the treemodel [02:34] You know what both python & mono need? The ability to just dump the VM state so you can debug at your leisure. [02:34] and spread around [02:34] in C, I'd have to write code to walk the linked lists and so forth [02:34] Nafai, can you not do a get_model() and then iterate on the roads in the debugger? [02:34] RAOF: That would be cool and I can't imagine why that shouldn't be possible with Python (not knowing the full details of the underlying python vm) [02:36] rickspencer3: I'm looking at the GtkMenu, not the GtkTreeModel. the GtkMenu stores the items in a glist, IIRC [02:36] which is a bunch of void pointers [02:36] so I have to walk the list, cast to the right object type, get the label property to see what I want [02:37] well, you could at least verify that the count is correct as a start [02:38] good point [02:38] :) [02:39] * Nafai is getting better at this with all your help [02:39] :) [02:41] heh [02:49] anyone here familiar with orbit? [03:02] looks like my HD pitivi movie is going to take about an hour or two to render :0 [03:02] * rickspencer3 goes to clean kitchen [03:02] chrisccoulson: My guess is that is a no :) [03:02] hehe [03:02] well, it's late [03:03] Nafai - i guess so. i don't think anybody understands orbit ;) [03:03] pitivi has gotten really good [03:03] chrisccoulson: isn't it eventually going away? [03:04] Nafai - yeah, gconf and bonobo are the big users [03:05] CORBA [03:05] * rickspencer3 shudders [03:05] heh [03:05] wsdl [03:05] soap [03:05] * rickspencer3 pukes enterprise alphabet soup [03:06] yeah, ewww [06:46] Good morning [06:59] pitti: Good morning to you. [07:00] hey TheMuso [07:00] * pitti -> breakfast === ogra_ is now known as ogra [07:53] good morning [08:08] morning [08:09] huats, hello [08:10] hey robert_ancell [08:10] how are you ? [08:11] huats, good, yourself? better luck with the package updates? [08:11] also, you going to uds this year? [08:12] robert_ancell, nope, I haven't been sponsored [08:12] huats, :( [08:12] AAArgh. Gosh darn it evolution. Is it *really* necessary for you to to use gigabytes of my ram *and* half my swap? [08:13] but actually I don't deserve really it since I haven't been very very active for a few months (do you know I am a proud and lucky father of a almost 3 months old little boy) [08:13] Congratulations! [08:13] thanks RAOF [08:14] that explains a lot my smaller contributions :) [08:15] (they will get bigger ;) ) [08:15] huats: ooh, congratulations! [08:16] thanks pitti :) [08:16] Theo + the building/running of my own company has been really really time consuming... [08:16] (I am not saying it is not worth it) [08:20] good morning [08:22] bonjour didrocks [08:23] Good morning didrocks, pitti. [08:23] hey RAOF, good evening [08:24] hey pitti, RAOF [08:24] pitti: how was your dinner yesterday? [08:25] Almost time to turn off all this computer equipment and dismantle the desk for moving! [08:27] RAOF: sure, enjoy your evening! === DrPepperKid is now known as MacSlow [08:36] didrocks: oh, wasn't dinner; my parents came by for congratulating, and we had some sparkling wine :) [08:37] ArneGoetje: could you please have a look at bug #563605 [08:37] Launchpad bug 563605 in language-pack-zh-hant "file overwrite error on upgrade" [Undecided,New] https://launchpad.net/bugs/563605 [08:38] pitti: sweet :) you move this week-end to Munich? [08:38] didrocks: no, next one, for a week; then further on to Belgium for two weeks [08:39] hum, right, it will be long for your family I guess… [08:39] grrr, I have this stupid applet bug display again [08:40] hey didrocks [08:40] didrocks: I'll come back to Dresden afterwards [08:40] hey pitti [08:40] didrocks: I'll probably spend about a week every month in Munich [08:40] bonjour seb128 [08:40] good morning seb128 [08:41] pitti: oh, moving back and forth. Hope it won't be too tiring for you [08:45] didrocks: should be fine; trains are comfortable :) [08:49] mvo: looking... [08:53] pitti, technically bug #550195 is not fixed, the issue is still in the mixer dialog [08:53] Launchpad bug 550195 in gnome-media "when volume is on 0 and moved up the volume doesn't update" [Low,Fix released] https://launchpad.net/bugs/550195 [08:53] seb128: hm; it seems to behave just fine for me [08:54] but we can reopen it if you still see something broken [08:54] pitti, oh, indeed [08:54] pitti, it was fixed in 2.29.91-0ubuntu1 [08:54] + bug 599663 - make changing the volume unmute (lp: #460390) [08:54] Error: Launchpad bug 599663 could not be found [08:54] Launchpad bug 460390 in gnome-media "Pulling volume down to bottom and up again leaves mute on" [Low,Fix released] https://launchpad.net/bugs/460390 [08:55] pitti, thanks for closing it ;-) [08:55] ah, great [08:58] pitti, so it seems that gdm bug which didn't make any sense about login failing is a nvidia binary driver bug [09:04] seb128: bug 516520 you mean? [09:04] Launchpad bug 516520 in gdm "pam_unix(gdm:auth): conversation failed" [Undecided,Incomplete] https://launchpad.net/bugs/516520 [09:04] seb128: funny, I was just going to look at it again [09:04] I'm done with my current task and am about to catch up on bug mail again [09:05] seb128: oh, it seems to be the reason for some people, but not others [09:07] oh, seems it was confirmed to be an nvidia issue indeed [09:07] * pitti dupes [09:08] pitti, right, the guy who was adding most of the recent comments said it's a nvidia issue for him tto [09:08] too [09:09] it's annoying when a bug is hijacked by n users having different issues [09:10] yeah, I asked to file new and separate reports for anyone not using nvidia 96 [09:16] mvo: rebuilding langpacks [09:16] ArneGoetje: thanks [09:17] mvo: the gimp2.0 templates have moved from -gnome to the common langpacks... I have changed that by adding an override. [09:18] * mvo nods [09:22] \o/ [09:22] I managed to go over 300 bugs closed for lucid ;-) [09:23] congrats seb128 :-) [09:23] didrocks, thanks === oubiwann` is now known as oubiwann [09:24] seb128: where did you see that? http://qa.ubuntu.com/reports/bug-fixing/lucid-fixes-report.html told 296 [09:25] didrocks, the url you were not listed on yesterday [09:25] the team one [09:25] I think the one you just looked at is one run behind [09:25] * didrocks doesn't like the team one. I don't exist :) [09:26] didrocks, I'm sure I can fix some other 5 by next week to be over 300 on the other one anyway ;-) [09:26] seb128: heh :-) [09:27] well, 117 on the non updated one. Not so bad for a beginner :p [09:27] indeed [09:27] though I would not call you beginner [09:27] well, let's say time parted and full time now :) [09:45] interesting, my rhythmbox process has a virt size of 1.1GB [09:45] it's just under 2 days old [09:47] what the what, why does it have libflashplayer.so mapped into its process?! [09:48] Ng, the music store [09:48] oh right [09:48] Ng, it's using webkit which loads everything a browser would load I think [09:48] yeah [09:48] fair point [09:48] still, 1.1GB virt seems somewhat unreasonable [09:48] indeed [09:49] those bugs are not nice to debug though [09:49] out of letting rhythmbox running under valgrind for 2 days [09:49] which I'm not even sure would work [09:49] valgrind is often too slow to get some usable [09:50] yeah [09:51] pmap doesn't show anything particularly outrageous, just lots of little anon allocations [09:55] it should probably disable the plugins unless it really needs them [09:55] the webkitview support that easily [09:57] I guess software-center does it? [09:57] yes [09:57] I can do a patch for u1 if rodrigo_ wants it [09:58] * mvo does one anywa [09:58] y [09:58] mvo, a patch for doing what? [09:59] rodrigo_, hey [09:59] rodrigo_, do you know if the music store needs flash in webkit? [09:59] Ng, ugh, 1.1GB is a lot [09:59] rodrigo_: disaable plugins: http://paste.ubuntu.com/414827/ should really be all that is needed [09:59] seb128, no, it doesn't, but as you say, it loads all the plugins, not sure how to make it not load them [10:00] (unless you need them for some reason) [10:00] rodrigo_, that's where mvo kicks in [10:00] mvo, ah, cool! [10:00] mvo, submit a branch and I'll approve/merge it [10:00] rodrigo_: I was about to say "I only have 3 albums in rb and only listened to one album of previews from U1MS since I started it", but then I remembered my iPhone which has ~30GB of music on it, so I'm guessing it's the plugin(s) for that that's to blame ;) [10:01] mvo, although, I guess we want the patch in the package, rather than waiting for a new release, right? [10:01] Ng, is your ipod connected? does the virt goes down if you disconnect it? [10:01] mvo, well, submit the branch anyway, and will talk with kenvandine so that he submits the patch [10:01] rodrigo_: I don't know the way your team works, the packages does not seem to have a Vcs-Bzr header so I need to go to the web to look where to find the branch [10:02] Ng, hmm, as seb128 says, try disconnecting the ipod? [10:02] mvo, it's using the canonical location [10:02] rodrigo_: ok, I can also just upload, whatever you prefer [10:02] seb128: aha, cool [10:02] seb128: I quit RB so my machine would stop swapping so much I couldn't type, but I'll start it up again and try an experiment [10:02] * mvo will push a branch [10:02] mvo, lp:libubuntuone is the trunk [10:02] Ng, thanks [10:02] thanks rodrigo_ [10:02] mvo, ok, as you prefer, but send me the patch so that I merge it into trunk [10:02] sure thing [10:03] ok, fresh rhythmbox starts with a 700MB virt, 70MB res - I assume that's because this is my first run with the new RB that shows U1MS by default [10:03] Ng, what is u1ms? [10:04] delta of 1 micro second? [10:04] just kidding [10:04] seb128: the music store [10:04] not enough sleep is not good for my humor [10:04] Ng, it uses 226M and 56M there [10:04] mvo, lol [10:04] seb128: 32bit vs 64bit maybe? [10:04] (I'm running 64bit) [10:04] 32 there [10:04] :) [10:04] and I've 1500 songs in my library [10:10] seb128: canonical location as in lp:ubuntu/lucid/libubuntuone ? [10:11] mvo, lp:ubuntu/libubuntuone [10:11] mvo, it's supposed to have the serie listed? [10:13] hmm, even after several attach/remove ipod events it's holding around 740MB virt 90MB res [10:13] seb128: I guess not, I'm just not using it that often yet, mostly using the old ~ubuntu-core-dev native branches [10:13] mvo, ok [10:14] what if you disable the U1ms plugin and start again? how much memory it uses when you first startup? [10:15] rodrigo_: 500MB virt, 43MB res [10:16] Ng, and if you enable it and start again? [10:16] ah, 700mb, 70, right? [10:17] rodrigo_: roughly, just did that and it's 644/64 this time [10:17] the music store plugin does not use much memory, just what the webkit widget does, so I guess most of that is the plugins [10:18] would be interesting to know how much difference mvo's patch makes there [10:19] yes [10:20] I have 169M 44640 26716 with the patch [10:20] lets try without [10:21] seb128: are you aware of bug 551809? [10:21] Launchpad bug 551809 in gnome-settings-daemon "gnome-settings-daemon crashed with SIGSEGV" [High,Triaged] https://launchpad.net/bugs/551809 [10:21] looks quite severe to me, with 50+ people experiencing it [10:21] no response from upstream [10:22] and 161M 45100 26716 without, that does not make much sense [10:22] * mvo looks into the details [10:23] milanbv, hey, not aware of it no, but a valgrind log would be useful [10:24] seems most users don't notice any issue out the crash icon displayed [10:24] so it's not really a concern for lucid [10:26] milanbv, the indicator-applet should be added on karmic upgrade btw [10:26] to reply to your bug comment [10:26] seb128: I notice the crash because nautilus doesn't get the theme right after that [10:26] I pushed the branch, I have a call now, and can look into it again afterwards only (https://code.edge.launchpad.net/~mvo/libubuntuone/webkit-no-plugins) [10:26] not sure that happens to other users, seems a bug in nautilus [10:27] we did add it back for any upgrade to lucid because users removed the indicator-message one without knowing it would contain other icons [10:27] but a valgrind log... I don't know how to reproduce the crash! [10:27] that's usually what you get when g-s-d stops running [10:27] theme got unset [10:27] and starting g-s-d back has some weird effects [10:27] yeah, but while most apps get it right, nautilus doesn't [10:27] ie some softwares don't update the theme [10:27] well that's another issue [10:28] right, that's a known one [10:28] for the indicator, I'm going to close the bug again then [10:28] nautilus is not the only one to not get the theme right [10:28] we'll see [10:28] thanks [10:28] so I should try to run it in valgrind an wait... [10:28] I guess those people didn't get indicator-sound installed for some reason [10:28] or upgraded before we had the code to add the applet to the config [10:28] right [10:28] seems it's happening on upgrade though [10:29] doesn't seem, but let's wait and see if more people come ;-) [10:29] so I would say maybe on icon theme cache update [10:29] try running gtk-icon-cache-update [10:29] and see if g-s-d gets valgrind errors [10:29] or try changing the themes [10:29] chrisccoulson: hey [10:29] how are you? [10:30] OK [10:30] hey seb128, i'm good thanks. how are you? [10:30] I'm good thanks ;-) [10:34] hey chrisccoulson, good morning! [10:34] chrisccoulson: sorry for my sudden disappearance last night; I wasn't sure when/whether my parents would come for the little congratulation/drinking session, and just when you asked they rang :) [10:34] hey pitti, how are you? [10:34] chrisccoulson: I'm fine doing the removals etc. now [10:34] good morning chrisccoulson :) [10:34] chrisccoulson: I'm great, thanks! how about you? [10:35] pitti - i'm good thanks. did i tell you which packages need removing? [10:35] chrisccoulson: I don't think so [10:36] pitti - stumbleupon (license issues) and livehttpheaders (that needs updating and it was always borderline whether it should stay anyway) [10:37] pitti - also, there may be some other removals for packages that haven't been ported from xul191 yet, as we still intend to only ship 1 xulrunner [10:37] chrisccoulson: in dubio pro abitus :) [10:38] where is the wiki page that explains how a gui program can add apport support? [10:40] pitti - the other likely removal is sugar. it currently has a depends on pyxpcom which used to be provided by xul191, but is now distributed separately by mozilla [10:41] and we haven't packaged that yet, and it's never had a proper release either [10:42] chrisccoulson: I propose that you open a bug (or find an existing one) about that, subscribe ubuntu-archive and me, and just keep adding tasks for packages to be removed [10:42] heh, good job i'm not flying today! [10:42] pitti - yeah, i can do that [10:42] chrisccoulson: then we have both a papertrail and also a more solid/distributed way of tracking this [10:43] seb128: would you happen to know why we suddenly have the empathy icon in the task bar again, in addition to the msg indicator? [10:43] this is a fairly recent regression [10:43] pitti, taskbar? [10:43] panel [10:43] I didn't notice that no [10:43] dunno [10:49] pitti - you're not autostarting empathy are you? [10:49] no [10:50] it happens on the mini and on my desktop now [10:50] hm, fun; restarting empathy again, and it's gone [10:50] maybe it was fixed with yesterday's updates, or it's another weird race [10:57] mvo, I'll propose your branch for merging, will be testing it later, since the music store web is down now [11:01] seb128: would you be interessted to test my debs for the disable plugins? I can upload them to people then [11:01] rodrigo_: I prepared a upload, but I would like some confirmation first that it actually helps :) [11:01] mvo, sure, i386 [11:02] mvo, wait then, the music store is down right now, so can't fully confirm it works [11:02] mvo, I'd say it's ok, but let's make sure better :) [11:02] ok [11:02] rodrigo_, the store is working there [11:02] caution++ [11:02] rodrigo_, at least I can browse it [11:02] seb128, hmm [11:02] I get a 503 error [11:03] but the store might be not the same since we are not the same country [11:03] +in [11:03] seb128, you get the EU store, as I do [11:03] k, works for me... [11:03] http://people.canonical.com/~mvo/libubuntuone/ [11:04] unless it thinks you're in .de :) [11:04] where do you think I am? ;-) [11:04] Germany and UK have their own stores, the rest of the EU gets the EU store [11:04] ask mvo :p [11:04] seb128, oh, where are you? [11:04] haha [11:04] seb128 is definetely german! [11:05] +1 [11:05] I would say France but apparently germans think it's close enough from Germany that it can be considered half German [11:05] also he *claims* to live in .fr [11:05] :) [11:05] seb128, right, so probably geoip also thinks you're german :D [11:05] see ;-) [11:05] * seb128 hugs mvo [11:05] * mvo hugs seb128 [11:05] :) [11:05] * mvo is curious for the result [11:06] seb128, what does it say on the home page of the store? something like 'EU store' ? [11:06] hey robert_ancell [11:07] robert_ancell, "EU Store Top Picks" [11:07] ups [11:07] rodrigo_, ^ [11:07] so yes it's the EU store [11:08] seb128, EU store? [11:09]