[00:26] alright, I fixed the bug.. Can someone help me figure out how to submit it? I tried submitting a merge proposal but that didn't work out right https://code.launchpad.net/~sethj/ubuntu/wily/unity/fix-for-1445595/+merge/276072 [00:28] quarter million lines changed, heh [00:30] sarnold, yeah. that was strange. I used bzr branch lp:unity to get the source.. [00:30] should I have instead pushed it to unity/unity instead of wily/unity? [00:31] sethj: sorry, I don't know bzr very well :/ [00:31] perhaps paste this link into the bug? http://bazaar.launchpad.net/~sethj/ubuntu/wily/unity/fix-for-1445595/revision/4028 [00:38] sarnold, I doubt it is as simple as that.. Somehow I got an old source tree so I'm "reverting" a bunch of changes. [01:22] sethj: this would happen if you starting with lp:unity and tried to merge to lp:ubuntu/unity [01:23] cyphermox, I tried merging into wily/unity, does that count? [01:23] same thing :) [01:23] that's what I thought. [01:23] run bzr info, check what the parent is [01:23] Should I remerge into unity/unity then? [01:24] parent branch is parent branch: bzr+ssh://bazaar.launchpad.net/+branch/unity/ cyphermox [01:24] right [01:24] so you simply want to file your merge proposal again, but against ~unity-team/unity/trunk [01:25] (lp:unity, should be in the list of suggestions I think) [01:26] you should be able to change that by using the resubmit link at the top right on the merge proposal page [01:27] thanks cyphermox! It looks like that worked. [01:32] sethj: yep. I'll let the unity dev review that === cute is now known as EphraimB [06:01] hi [06:35] Morning hikiko [06:37] hi duflu :) [06:37] how are you? [06:37] hikiko: Going OK. How are you? [06:42] I'm good too [06:54] seb128, I gotta run, but tried to push a no change rebuild of g-c-c for cheese, but it needs the new valac package MIR'ed [06:54] seb128, also unity is FTBFS with what looks like gcc5/libsig++ fallout, https://launchpad.net/~darkxst/+archive/ubuntu/gd318/+build/8216566/+files/buildlog_ubuntu-xenial-amd64.unity_7.3.2%2B15.10.20151016-0ubuntu2_BUILDING.txt.gz [06:58] good morning desktopers [06:59] hey duflu hikiko darkxst [06:59] darkxst, Trevinho has unity fixed from what he said in the previous days [06:59] k for vala [06:59] good morning [07:00] Morning seb128 [07:00] And morning didrocks [07:00] lut didrocks [07:01] hey duflu, seb128 [07:02] hey seb128 didrocks [07:02] salut hikiko [07:23] Good morning [07:24] good morning didrocks, duflu, seb128, hikiko, pitti, darkxst :) [07:24] (and everyone else) [07:24] hey desrt, wie gehts? [07:24] hey desrt pitti [07:24] Hey desrt [07:24] pitti: i gehts very well, thank you [07:24] goodmorning all :p [07:25] * seb128 goes to get something to eat, bbiab [07:26] morning pitti, desrt [07:31] greetings, mvo [07:31] hey desrt [07:33] ça va seb128 et didrocks ! [07:33] hey mvo, wie gehts? [07:34] hey pitti, I'm very well, thank you! und du? [07:34] mvo: bisschen muede, bin erst halb zwei ins Bett; aber okay, danke! [07:35] pitti: halb-zwei? hat augsburg noch gespielt letzte nacht :) ? [07:36] mvo: Actually FCA did play yesterday evening for the DFB Pokal (3:0 against Freiburg!), but that wasn't the reason [07:37] pitti: sneezing way more and didn't really sleep (was awake between 3:30am to 7am), so well… :p [07:37] mvo: I play basketball Wed evenings, and only come home 22:30, and after that I'm still too pumped to go to bed immediately [07:37] so, late-night debugging session :) [07:37] didrocks: ah.. you too, hm? [07:37] * desrt is also coming down with something and didn't sleep well :( [07:37] desrt: yeah, since Monday, just getting worse [07:38] ugh [07:38] i was sort of hoping it cleared up before i have to fly [07:39] pitti: aha, nice! [07:39] didrocks: erk; time for some tea and midday nap? [07:41] pitti: good idea, and will add some honey to the tea as well! [07:42] * desrt makes some tea too [07:51] hey desktopers [07:52] could someone hold back gjs to transition from xenial-proposed? [07:52] afaics it breaks API/ABI [07:53] ricotz: it still builds libgjs0e, so if it does change ABI it doesn't declare it? [07:54] ricotz: you can file a bug against it and tag it "block-proposed"; right now it's held because of the powerpc FTBFS, but of course that might get "accidentally" resolved; so better file a bug with a rationale [07:54] pitti, upstream doesnt care about abi break, that is why there is the suffix "e" [07:56] pitti, ok [08:02] pitti, is this enough since I can't link it xenial -- bug 1511253 [08:02] bug 1511253 in gjs (Ubuntu) "1.44.0 breaks API/ABI without transitioning" [Undecided,New] https://launchpad.net/bugs/1511253 [08:02] ricotz: yes, proposed-migration doesn't care about that; but I'll mark it anyway for clarity [08:03] thx [08:06] good morning! [08:08] Morning! [08:12] larsu: hello! [08:13] morning larsu & Trevinho [08:13] hi Trevinho and desrt! How are you? [08:13] and didrocks :) [08:13] Trevinho: hey :) [08:13] not great. sick :/ [08:13] didrocks gave it to me [08:14] :( [08:14] get better! [08:15] Hey larsu and didrocks! So, so... You? [08:17] sick as well :/ [08:17] missing quite some hours of sleep in the process due to this [08:17] Trevinho: good good. Going to Milan this afternoon [08:18] didrocks: uh oh - got worse after all? Sorry to hear [08:18] larsu: yeah, sneezing and couldn't sleep anymore between 3:30 to 7am [08:24] larsu: Milan? Why? [08:26] Trevinho: Faina has some work thing there. I'm coming as the spouse [08:26] and then we'll spend some time there [08:26] drinking a lot of coffee, I presume :) [08:31] larsu: eh he... Good, it's a good moment for Milan, because of the expo... [08:31] I guess you'll find a lot of other good stuff, together with espresso :-) [08:33] hehe [08:33] * larsu is looking forward to it [08:39] desrt: g_settings_set() immediately changes the value in the settings object and sends a changed() to the same process? [08:40] yes [08:40] meanwhile it sends a dbus message (async via the worker thread) to dconf to record the change [08:40] right [08:40] just making sure [08:40] when dconf says "okay, i'm done" then the local value is dropped from the process [08:40] having some trouble with the qml bindings again [08:40] or if dconf says "something went wrong" then the local value is dropped and an additional change signal is sent (so that the program sees the old value again) [08:40] desrt: oh you *remember* the local value? [08:40] yes [08:40] we have to, since it's not on disk yet and the user will expect to see it immediately [08:41] makes sense [08:41] dconf is very tricky [08:41] so is irccloud... [08:41] why? [08:41] i just realised while having this conversation that it does much the same thing [08:41] when i press enter, it immediately shows my message as having been sent [08:41] seb128: are you on teh remaining bits of the poppler transition? (http://people.canonical.com/~ubuntu-archive/proposed-migration/update_output.txt) [08:41] hehe [08:41] but by the time it makes it to the server, maybe you said something before i did [08:41] seb128: seems there's still some uninstallables there [08:41] so it reorders messages after they've already been put in the buffer [08:42] pitti, I plan to spend some time today looking at some of the transitions, cheese poppler evolution-data-server [08:42] but help would be welcome [08:42] cool, thanks [08:42] desrt: the problem is that we're now sending the changed signal in another main loop iteration (because of a qt bug). When the change signal comes in, we don't emit it anymore because we have a "if the value hasn't change, don't emit" condition [08:42] and I can't take that out, becasue of cycles [08:42] it's all very messed up [08:43] pitti, let me do a round of test builds now [08:43] ugh [08:44] gsettings handles this in a pretty simple way [08:44] seb128: it's just libpoppler56 -> 58, -glib, -qt etc. are the same, right? [08:44] seb128: want a transition tracker for it? [08:44] 1) change signals that came from the same process always go directly, never via mainloop [08:44] pitti, correct [08:44] s/process/thread/ [08:44] pitti, that could be good yes [08:44] 2) we detect recursion in the normal sense of the word and stop it there [08:44] desrt: (1) is what we had before, but we have to go through the main loop because qt leaks if we don't [08:44] pitti, we might want one for e-d-s as well? [08:44] good luck to you... [08:45] i don't know of a good solution to this problem. this problem is part of the reason that gsettings works the way that it does. [08:45] desrt: I could now remember the changed keys in a hash map and check that when change signals come in [08:45] or remove the check entirely [08:45] I hate this problem :/ [08:51] pitti, some of the rebuilds are failing due to valac not being installable if you want to have a look to that [08:51] seb128: won't have time today, I'm afraid [08:51] k, no worry [08:51] adding to my todo if nobody beats me [08:52] +to it [08:53] seb128: committed poppler transition [08:54] pitti, danke [09:01] morning all [09:02] hey willcooke [09:03] hey willcooke [09:04] hello willcooke [09:04] morning willcooke [09:23] larsu, I think it was you who told me about the calendar trick where you set the timezone to ???? somewhere in Iceland perhaps, and then it doesn't move around with DST changes? [09:23] Can you remind me where that place was? [09:24] oh wait, ignore [09:24] I've just removed the time zone altogether [09:28] willcooke, Laney does that ;-) [09:29] ah, thx [09:29] I think it's sorted now [09:42] * larsu has no clue what willcooke is talking about [09:42] * willcooke puts down the crack pipe [09:43] hehe [09:57] https://bugs.launchpad.net/ubuntu/+source/nautilus/+bug/1480217 [09:57] Launchpad bug 1480217 in nautilus (Ubuntu) "Nautilus background handling screwed when changing scaling factor." [High,Triaged] [09:58] We've been asked to take a look at this bug with some urgency, who can take a look? It'll need back-porting to Trusty as well [09:59] larsu, ^ maybe? [10:00] Laney and larsu are the only ones in the team with hidpi configs I think [10:00] reading right now [10:00] danke [10:00] thanks larsu [10:13] damned you arm64! [10:13] Unity doesn't compile there (in xenial). collect2: fatal error: ld terminated with signal 11 [Segmentation fault] [10:13] Not much to do... Nor to understand [10:13] just blame doko and move on :) [10:24] Trevinho, try to ping doko about it on #ubuntu-devel [10:26] seb128: ok thanks === nudtrobert1 is now known as nudtrobert [11:01] Can we get this SRU'd in the next few days? https://bugs.launchpad.net/ubuntu/+source/gtk+3.0/+bug/1508357 [11:01] Launchpad bug 1508357 in gtk+3.0 (Ubuntu) "Tooltips have a black squares outside of its rounded corners" [Medium,Fix committed] [11:01] willcooke, yeah, it's on my list for today [11:02] oh sweet, thanks seb128 [11:02] willcooke, did we get any user reporting it or commenting about the issue? [11:02] I didn't see any comment about that one yet [11:02] seb128, not many. I saw a guy mention it on the Ubuntu G+ community, but I dont think people are really noticing it. [11:03] didrocks has good eyes ;-) [11:03] :) [11:03] I still can't see it even on my external monitor [11:03] for close vision, sure :) [11:03] well I see that the corners are not round [11:03] same [11:03] but I can't see the round inside the rectangle [11:03] * larsu sings the song of rgba windows [11:04] seb128, I can just about make it out on my Inspiron [11:04] but I have to press my nose against the screen :) [11:04] :-) [11:04] willcooke: touch screen? [11:06] :D [11:49] Sweet5hark1, please ping me when you copied 5.0.3~rc2 to the public ppa [11:51] ricotz: aye [11:52] Sweet5hark1, and don't be lazy with tweaking the version for a ppa ;) [11:53] Sweet5hark1, while you are here, do you see a need for 5.1.x backports for vivid? [11:57] ricotz: not really. vivid is EOL when 5.1.0 will be released. [12:00] ricotz: 5.0.3 copied. [12:09] Sweet5hark1, alright [12:11] Sweet5hark1, do you intend to push a xenial build for 5.0.3 to the ppa too? [12:14] good morning [12:15] qengho: hello! [12:16] popey: FYI, the twine support brings interesting challenge for ubuntu make :p [12:16] Oh good! :) [12:16] node? [12:17] "challenge" [12:17] Sweet5hark1, ooh, do you have a 5.0.3 build for me too? :) [12:17] popey: not, more simple than that: they don't provide an icon, so I need to do multiple downloads in parallel [12:17] hah [12:17] popey: in addition to the requirements (packages to install) [12:17] it was supposed to be supported from the beginning [12:17] but it seems there are BUGS! :) [12:18] (like a callback only called once called multiple times…) [12:18] weird that unit tests are still telling it's doing the right thing, need to investigate… [12:18] tkamppeter, what's the difference between Debian's pysmbc and our python-smbc? should be use the same name as they do? also it seems there is a new bugfix update available if you want to work on that one [12:20] didrocks: welcome to TDD. the API is what the tests say it is. please adjust your code to be more like the tests :) [12:24] desrt: actually, the test is correct and the module have the right behavior [12:25] the issue is in the main module doing the different calls not behaving well [12:25] (the icon file is too small and so one download finish before I get the chance to have one progress for all requested download) [12:37] seb128, thank you very much. I was not aware of that. I will have a look. By the way, what is the (Ubuntu/Debian) naming policy for python libs, python-xxx or pyxxx? When I started at Ubuntu there were simply these python-xxx packages for s-c-p which I continued to maintain. [12:37] tkamppeter, unsure there is a policy about the source packages names, usually we use the same as upstream does [12:39] yeah, for instance python-foo would be "foo" as source package name (for the most of python ones I've seen) === hikiko-lpt is now known as hikiko|ln === alan_g is now known as alan_g|lunch === hikiko|ln is now known as hikiko === alan_g|lunch is now known as alan_g [14:05] oh nice, CI jenkins now seems to have access to docker hub [14:05] * didrocks will be able to wire medium tests [14:05] neat! [14:06] * didrocks takes his plumbing hat, and start diving [14:08] * mdeslaur slaps seb128 for losing changelog entries in unzip merge [14:08] mdeslaur, ? [14:08] oh, you mean merging back old changelog entries? [14:08] yeah [14:08] oh, did it get synced? [14:09] question for you, if we had unzip in sync for a week and then added a delta again, would you then bring back all the old entries? [14:09] no [14:10] k, so if that makes you happier just think that it was synced by mistake [14:10] * mdeslaur slaps seb128 again [14:10] then some change needed were added :p [14:10] roh [14:10] ahah, that discussion again! [14:10] but yeah, I can't be bothered [14:10] slapping you makes me happier :) [14:10] * didrocks has some déjà-vu from 2009 :) [14:10] I never merged the old changelog entries, that's more work than it's worth and it eats extra space on disk [14:10] didrocks, yeah ;-) [14:11] mdeslaur, instead of slapping me you should get going with doing your merges! ;-) [14:11] that wouldn't happen if you were the one who merged it :p [14:12] hehe :) [14:26] willcooke: welcome back and ping about UOS sessions from the desktop team for next week [14:26] mhall119, hi, going in to a meeting in a mo [14:26] willcooke: let me know when you're free [14:28] that reminds me, seb128 can you attend: http://summit.ubuntu.com/uos-1511/meeting/22576/developer-desktop-plan-1604/ [14:28] willcooke, what tz is that in? utc? [14:28] seb128, I think so [14:29] so yes [14:29] thx [14:30] yes, it's UTC [14:38] mhall119, meeting done quickly. [14:38] mhall119, so fire away [15:02] hey willcooke, I was asking for UOS sessions for next week, we have the developer desktop planning on already, but are there any others from your team that we can put on the schedule? [15:02] mhall119, we've added a QA one too [15:03] ok, any others you can think of? [15:03] other than that, there isn't anything else we need to meet about, the general 16.04 will cover everything we need to cover [15:03] ok [15:05] seb128, I have checked now and Debian has pysmbc 1.0.15.4-1 with binary package names python-smbc and python3-smbc, we have python-smbc 1.0.15.4-0ubuntu1 with the same binary package names. [15:06] seb128, how should I proceed to sync Debian's packages? Simply syncpackage pysmbc? Or do I (or the Debian package maintainer) do anything additional, for example to get the old python-smbc source package removed? [15:14] tkamppeter, the source needs to be removed from http://people.canonical.com/~ubuntu-archive/sync-blacklist.txt [15:14] then we need to remove the ubuntu specific source [15:18] seb128, can I already syncpackage the new ones? The version number of Debian is "newer" so with both the python-smbc source package from Ubuntu and the pysmbc source package from Debian in our archive the Debian binary packages should be picked by installation processes. [15:18] tkamppeter, you can try [15:20] seb128, the sync is blocked by the blacklist. Who can remove the entry? [15:21] tkamppeter, I probably can, let me have a look [15:21] or I see that you pinged Colin [15:21] seb128, if you can, simply do. [15:23] tkamppeter, done [15:24] seb128, thanks. [15:24] yw [15:25] seb128, how long does it take for really getting active? [15:25] that I don't know [15:25] Colin should know though [15:29] seb128, now pysmbc is introduced as a new source package probably going into Universe, but the binary python3-smbc is in Main and the new one should continue in Main. What needs to be done here. [15:33] tkamppeter, I've accepted it to main [15:33] seb128, Thanks. [15:35] seb128, and how will the old, now useless, python-smbc source package be removed? Can you simply remove it? [15:35] yes, going to do that once the new is built [15:54] Sweet5hark1, do you intend to push a xenial build for 5.0.3 to the ppa too?? [16:01] ricotz: not really [16:01] ricotz: will work on 5.1 alpha for xenial [16:01] Sweet5hark1, hmm, will do it then [16:01] (to cope with python3.5 and poppler) [16:02] Sweet5hark1, see pm === davidcalle is now known as davidcalle|afk [16:41] Trevinho, got my timezones all mixed up so I'm not sure exactly when morning is for you, but I found the issue behind the empty trash bug. [16:41] sethj: hey [16:41] and I *think* I fixed it. [16:43] I mean, I fixed it and everything seems to work still haha. it was pretty simple so there shouldn't be any room for breaking something else. [16:44] sethj: I've seen the change... [16:44] The ... doesn't make me feel good D: [16:44] However that causes another issue: so basically the "empty trash" dialog might not be focused [16:45] i.e. if you write something in gnome-terminal or gedit, then you empty the trash, the dialog won't be focused [16:45] That's because we need to send the action timestamp... The way i'm doing it's hackish I think [16:45] sethj: the proper solution would be to fix the Nautilus dbus API to get the action timestamp as parameter. [16:46] sethj: another possibility instead is to do what you did, and... To change nautilus to force opening that dialog in front always [16:46] that can be done by using gtkx stuff, but it's a little hackish [16:46] I guess having proper API would be better [16:47] Trevinho, let me do some more testing on the focus issue. The dialog always raised in focus for me.. [16:48] That makes sense why the Activate call was there then. That was really confusing me (hence the suspicion that something was off with my fix) [16:48] sethj: the related bug was https://bugs.launchpad.net/ubuntu/+source/unity/+bug/807808 [16:48] Launchpad bug 807808 in Unity "Empty trash dialog sometimes dont have the focus" [Medium,Fix released] [16:49] Trevinho, cool. I'm no wizard, but I will take another look. [16:49] sethj: thanks for taking care of this issue :) [16:52] seb128, seems that the new pysmbc has correctly made it into the archive now. so the old python-smbc can get removed. [16:52] tkamppeter, k, thanks [16:53] ok, headache too strong, needs to take a nap, see you tomorrow guys [16:55] Trevinho, do you have any more precise steps to reproduce? The dialog always appears on top and in focus.. [16:57] sethj: you need to write in some gnome app, then right-click the trash icon and select empty the trash.. [16:57] Trevinho, that's what I'm doing... [16:57] sethj: maybe something changed at nautilus level nowadays, so it might be not neeeded... [16:58] Trevinho, so I tried slack, gedit and gnome-terminal. with text typed and the text fields in focus the dialog still takes focus and the buttons can be manipulated from the keyboard.. [16:59] sethj: you also have to ensure that no nautilus window was opened.. [17:00] Trevinho, nope. [17:00] sethj: basically, the same issue (the focus thing) happens if you go in the dash, files lens, open a folder preview -> press on "Open" button [17:01] That action doesn't use the proper activation (there's no timestamp in previews, yet) so... The nautilus window instead of being focused, it gets opened underneath your latest used app [17:02] sethj: but... In case this has changed recently (as per some nautilus change), then we could try your fix [17:02] Trevinho, I can't reproduce the dash issue either. [17:02] I might have noticed one regression though, testing.. [17:03] mhmh, sethj have you maybe disabled the compiz focus stealing prevention? [17:04] sethj: in ccsm -> general -> Focus and behavior -> Focus prevention level by default is "low" [17:04] Trevinho, ahh, yes, I believe I did. Why is that at low? It's so awful. [17:05] sethj: well, the thing is that if you're writing something... And something opens, the focus shouldn't be moved away [17:05] or you risk to continue writing, pressing keys in another window [17:06] ah, right. [17:06] Now that I think about it that happened to me during the Wily upgrade. [17:07] yup, if I put focus prevention on low the dialog doesn't raise with focus. [17:07] grr. [17:08] so... Yeah, the only way we have do do this cleanly is to add a timestamp parameter to nautilus API [17:09] This won't work when not using nautilus (i.e. elementary file manager used to work), but well... Not supporting ubuntu default is hard [17:10] honestly I wish the default was Thunar :P Okay, I'll see if I can fix it differently. [17:11] You can add an XUbuntuEmptyTrash method with timestamp parameter [17:11] and... if you want thundar to support it, it has to be patched as well [17:14] Trevinho, there's no real point IMO since Unity doesn't support changing the default file manager anyway.. [17:15] but I'll see what I can do with nautilus. I'm not sure I understand what makes the timestamp so crucial though.. [17:15] Mh, I see [17:16] sethj: basically you should use gtk_window_present_with_time... [17:17] sethj: now, there's a trick to get the current time from Xserver, and just use that value without passing it [17:17] but, the right way to do this, is to use the timestamp of the event that triggered the action [17:21] Trevinho, that's really helpful, but my first question was why adding a timestamp argument to the API fixes the issue. Does that somehow let it bypass the focus prevention? [17:43] sethj: yeah, basically there's a workaround to activate a gtk app by passing the timestamp via a gtk action... [17:44] sethj: it worked fine so I didn't change the API... But I didn't think to this sideeffect [17:44] ..... but please don't [17:44] Trevinho, ahh. [17:44] * larsu hopes will disallow apps from doing this [17:44] *mir [17:45] well mir will have to workout the other focus bugs then. [17:45] which other focus bugs? [17:56] larsu, the ones Trevinho has been working around. e.g. https://bugs.launchpad.net/ubuntu/+source/unity/+bug/807808 [17:56] Launchpad bug 807808 in Unity "Empty trash dialog sometimes dont have the focus" [Medium,Fix released] [17:58] Trevinho: wow! Why aren't we just fixing this in nautilus? [17:58] sethj: clearly the window manager shouldn't need to work around such issues [18:00] larsu: I do agree, but it needs an API change [18:00] dbus Api change, or DBus api addition [18:00] Trevinho: because it's an action activation? [18:01] doesn't that take platform data already? [18:01] nope [18:01] ah, this is still the old stuff with the command line? === alan_g is now known as alan_g|EOD [18:06] that might be fixed in newer nautilus versions? [18:08] seb128: I don't know if it has been converted to a dbus-activaeted app yet [18:09] I don't think it has [18:12] larsu, https://git.gnome.org/browse/nautilus/commit/?id=a5617016d3e5d8e32ce02708f2c08d09926141d1 [18:13] but unsure that's enough/what you mean [18:14] oh, that doesn't include the trash thing [18:14] is this hard coded in unity maybe? [18:26] larsu: yeah, for now we used an ugly workaround by http://pastebin.ubuntu.com/13001134/ [18:27] Trevinho, wow, so is the entire activate method just to "fix" the no focus bug? [18:27] sethj: yeah... [18:27] ouch. [18:36] * Trevinho leaves [19:31] hello everyone! quick question: I upgraded to Wily on my Lenovo Thinkpad X201 and the battery icon is missing, and so are the power settings [19:31] is there any package I should check is installeD? [19:32] indicator-power shows as installed at the newest version [19:52] g'night all - see you next week === davidcalle|afk is now known as davidcalle === howefield_afk is now known as howefield === howefield is now known as howefield_afk