[00:01] Sorry, xchat works (I tried to do something else before)... But still nothing for pidgin [00:01] Are you using the liferea from intrepid or hardy? [00:02] intrepid [00:02] but iirc, it was on hardy too [00:02] started a few months ago [00:02] Ok, I'll test both... [00:02] If it happened few months ago, then it is probably a bug in liferea :/ [00:05] Yep. Liferea in hardy shows the same behaviour as Pidgin... [00:06] maybe the notification area applet [00:06] I suppose... [00:06] To report a bug upstream? [00:07] (unless they say it's a feature :)) [00:07] i doubt it, it's a regression imho [00:07] I suppose they have some sort of BTS... [00:28] can't find anything related in launchpad [00:29] fta: This might be useful - http://sourceforge.net/tracker/?atid=581684&group_id=87005 [00:31] well, at least for reporting a new bug... it doesn't list any bug with "workspace" in summary [00:38] hmm, could be a gtk bug [00:38] Well, if one group of programs has one behaviour, and the other has different behaviour, then it might be the way programmers wrote the programs. [00:39] i'm reading liferea's code [00:39] I can help you with xchat's code in 10-20 minutes, if you want... [00:40] xchat is fine for me [00:41] It's fine for me too... We could try to look for the code that does on_notification_hide/show/click/whatever_is_its_name in both programs and compare the differences :) [00:48] ok, this is bogus [00:48] if((gdk_window_get_state(GTK_WIDGET(mainwindow)->window) & GDK_WINDOW_STATE_ICONIFIED) || !GTK_WIDGET_VISIBLE(mainwindow)) { [00:48] that's for showing window? [00:49] that's the test to see if the window must be hidden or made visible [00:51] Could you paste that piece of code to paste.ubuntu? I would like to see it [00:55] gdk_window_get_state(GTK_WIDGET(mainwindow)->window) & GDK_WINDOW_STATE_ICONIFIED returns 2 when the window is visible in another workspace [00:55] And it's recognized as true, if it's not 0 [00:56] and returns 0 when it's invisible (docked) or visible in the current worspace [00:56] Aha [00:56] !GTK_WIDGET_VISIBLE(mainwindow)) is fine [00:56] fta: Error: I am only a bot, please don't think I'm intelligent :) [00:56] lol [00:56] hehe [00:57] Should the first part check if it in another dock, or invisible? [00:58] *it's [00:58] *Shouldn't [00:58] http://paste.ubuntu.com/18200/ [00:59] ui_mainwindow_save_position() only saves X and Y values? [01:00] *saves only (note to self: learn to write) [01:00] http://paste.ubuntu.com/18201/ [01:00] thanks [01:01] Yep, x, y and w, h [01:01] and panes, but not workspace, it's window manager's business [01:02] if i iconify liferea in the current workspace, the 1st test returns 2 too :( [01:03] so visible in a different workspace = iconified for gtk [01:03] damn [01:07] damn, gdk_window_get_state(GTK_WIDGET(mainwindow)->window) returns 2 so it's not even a mask issue [01:08] http://library.gnome.org/devel/gdk/unstable/gdk-Event-Structures.html#GdkWindowState [01:09] yes, that confirms it [01:10] xchat may do something different [01:10] I'll take a look now [01:11] just to downlaod it [01:11] *download [01:24] fta: This is the code which is called when on window hide/show [01:24] http://paste.ubuntu.com/18207/ [01:24] for xchat [01:25] Do you thank that changing it to GTK_WIDGET_VISIBLE in liferea would do any good? [01:25] s/changing/using/ [01:25] s/it to // [01:26] That is, just removing the first check in if [01:26] leaving it to if(!GTK_WIDGET_VISIBLE(mainwindow)) [01:28] maybe. but the code is there for a reason. do we have a vcs to look for blame ? [01:29] Found one in debian [01:29] Vcs-Browser: http://git.debian.org/?p=collab-maint/liferea.git [01:29] not useful... [01:30] here's upstream: http://liferea.svn.sourceforge.net/viewvc/liferea/ [01:35] not enough history :( [01:36] but this also means that code is old, so something else caused the regression [01:37] Well, maybe there was the change to that part... the revision list for it is long [01:37] http://liferea.svn.sourceforge.net/viewvc/liferea/trunk/liferea/src/ui/ui_mainwindow.c?view=log [01:37] that function didn't change in 2.5y [01:37] or more [01:38] damn [01:41] indeed, that fixes it http://paste.ubuntu.com/18214/ [01:41] Woohoo, we rock :) [01:42] Are you gonna forward it upstream, to see if they want to include that change? :) [01:42] the only problem is that if you iconify it, you have to click twice to make it reappear [01:43] And now it's only once? [01:43] no, that's now [01:43] after the patch [01:43] that may be the purpose of the 1st test [01:44] Maybe... But, I don't see how it can be related to the mouse click [01:44] Maybe it's get_state function [01:45] eh? ui_mainwindow_toggle_visibility() is called after the mouse click is detected [01:45] Then again, the function should be called based on event that is caused by a click. [01:45] it is [01:45] Well, maybe we can just change it to react to a click, instead of a double-click [01:46] i said click twice, not double click ;) [01:46] Oooh... right :) [01:46] it's a status issue [01:47] So, if it's visible on another workspace, GTK_WIGDET_VISIBLE might return true, therefore going the the else-branch and hiding the window [01:47] yes [01:47] And then on second click it shows the window, since G_W_V is false [01:47] hence the GDK_WINDOW_STATE_ICONIFIED test [01:48] which is buggy [01:48] damn [01:48] there must be some other way. [01:48] Do you know if GTK is aware of current workspace? [01:48] and the workspace window is on? [01:48] no idea [01:49] at least we've cornered the bug [01:49] Right :) [01:50] Now just to find something that will do the "curWorkspace != windowWorkspace" check [01:50] :) [01:50] i'm not sure it's possible [01:50] Neither am I... As you said, it's WM's job. [01:53] Dunno if this check is worth it [01:53] GDK_WINDOW_STATE_WITHDRAWN the window is not shown. [01:54] the test returns only 0 or 2 [01:54] so GDK_WINDOW_STATE_ICONIFIED or nothing [01:55] But gtk_window_get_state can return any of the states (0, 1, 2, 4, ... ) [01:55] yes, but here, it's 0 or 2 [01:55] sorry :) [01:55] sniff [01:55] indeed [01:56] liferea makes little bunny cry [01:56] But, should the withdrawn state mean something like "it's not on this workspace, or it's minimized, or something similar" [01:56] ? [01:57] donno [01:58] Neither do I... Just guessing :) [02:06] http://ubuntuforums.org/showthread.php?t=821994 [02:07] heh ... [02:08] fta, in what IDE did you test liferea? [02:09] printf() [02:09] :) [02:09] Nice one :)... so, you edited the file, built it and then just watched the result? [02:09] yes [02:10] Ok, I just have to try that with withdrawn state ... it bugs me :) [02:10] fprintf(stderr, "something"); to be sure it's not buffered or logged somewhere in X [02:10] (even if it won't work) [02:10] thanks :) [02:12] fta, xchat works with two clicks too... [02:13] And they also test just GTK_WIDGET_VISIBLE [02:13] so that's expected [02:13] But, true... it would be good if it could just switch workspace with one click :) [02:47] fta: This is what I get from gtk_window_get_state: [02:47] gtk_w_g_s == 0gtk_w_g_s == 1gtk_w_g_s == 0gtk_w_g_s == 1gtk_w_g_s == 2gtk_w_g_s == 3gtk_w_g_s == 0 [02:48] for various clicks (depending on state and workspace) [02:50] and this is for output: fprintf(stderr, "gtk_w_g_s == %d", gdk_window_get_state(GTK_WIDGET(mainwindow)->window)); [03:02] fta: So, it seems it returns 0 when the window is on current workspace and we iconify it [03:03] 1 when we restore it and it was iconified it from the current workspace [03:03] 2 when the window is not on the current workspace, and we iconify it [03:03] and 3 when we restore it, and it wasn't iconified from the current workspace [03:04] Now I'm confused. How it returns this numbers, and GdkWindowState only knows for 0, 1, 2, 4, 8 (and 16) [03:05] Ok, let's change for 2, and 3 [03:05] for 2, is not on the current workspace, or it is minimised [03:06] same for 3, it also returns that if it was minimised [03:07] I'm gonna print all states, so to see which one is what number. I'm confused [03:26] fta: ok, all states are ok. 1, 2, 4, 8, 16, 32 and 64. But gtk_window_get_state() still returns 0... Why? === asac_ is now known as asac [12:09] asac: there? Maybe? :) [12:12] gnomefreak: There? [12:13] im here sort of [12:13] i cant open bug 1 and i really need to [12:14] gnomefreak: Error: Could not parse data returned by Launchpad: The read operation timed out (https://launchpad.net/bugs/1/+text) [12:14] ubottu: its ok [12:14] Factoid its ok not found [12:15] I think it's Launchpad... It is opening for me, but slowly [12:15] Regarding agenda items for the meeting. We shouldn't fill the "Decision" column before meeting. Can you move them to "Notes" column? I don't know where would you put them, so I didn't touch them. And I fixed that last agenda item, so now it's ok [12:16] gnomefreak: I was wondering if you know do we use #ubuntu-meeting or -mozillateam for the meeting? [12:17] And if we use the first channel, where do we "make a reservation for it" or something? [12:19] Jazzva: we will use -meeting but i will set that up this week' [12:19] Jazzva: i got that handled [12:19] i hate assholes [12:20] gnomefreak: Ok, I'll just send the announcment to the list and members. [12:20] Jazzva: hold off on that for now [12:20] i need to make sure noone has claimed that spot yet [12:20] Ok [12:20] i will do after i finish my email [12:32] Jazzva: ok wiki is fixed [12:32] k [12:32] next i will work on -meeting [12:32] give me some time on that since its sunday morning [12:32] Jazzva: wtf [12:33] Jazzva: i got rid of the last comment and the whole fucking thing comes undone [12:34] hmmmm [12:35] It's fixed. There was a space at the end of the line [12:35] Jazzva: can you please let me know how to fix that so i can stop it from happening [12:35] ah [12:35] you have got to be kidding me [12:35] Nope... that breaks the formatting :) [12:35] Jazzva: thanks [12:35] thats not fixed [12:35] now 2nd one is borkee [12:35] borkled [12:36] Really? Ok, I'll look again. It's still saving the current one. [12:36] thanks [12:36] np [12:36] i need smoke [12:37] Ok... where can we check ubuntu-meeting availability? [12:42] Jazzva: i have sunbird do that for me [12:42] Jazzva: fridge.ubuntu.com [12:42] Thanks :) [12:42] its clear [12:43] Can we reserve it for 22nd June, 20:00 UTC+2? [12:43] Jazzva: working on it [12:46] Sent the announcement... [12:46] Jazzva: i know a few people that edit the fridge and im trying them first there is also a email to request meetings with but #fridge is being moved to #ubuntu-news so im replying on editors i know since neither channel lists the email address [12:47] Jazzva: thanks. where? [12:47] list and members. I couldn't find some members' e-mails, though. [12:47] ah ok [12:49] [reed], if you read this the Mozilla team meeting is on 22nd June, 20:00 UTC+2 in #ubuntu-meeting. I couldn't find your e-mail, and I'm not sure if you're on the list [12:49] wait a minute please use UTC not UTC+2 :( [12:49] * gnomefreak writing email [12:50] so thats -600 for me? [12:50] Yes. [12:50] I'll edit the wiki page. [12:50] 1400 == 2pm [12:50] right [12:53] im asking cody i wish i didnt have to [12:53] i have email ready to send but he said i can ask there so lets see what he says [12:54] Ok, I changed the times on the wiki pages... [12:55] and i added the time link [12:56] wiki is slow today [12:56] :) [12:57] Jazzva: now when you click the time/date you get a conversion chart [12:57] :) once it saves [12:57] Right... but you set it for 3rd June, 9pm :) [12:57] http://fridge.ubuntu.com/node/1516 [12:58] Jazzva: i didnt change the date or time [12:58] i just added the link [12:58] When: [WWW] Sunday, 22nd June 2008, 18:00 UTC [12:58] looks goo to me [12:58] good [12:59] our meeting is set up for #ubuntu-meeting [12:59] Following the link: "Time is fixed on Tuesday, June 3, 2008 at 21:00:00 UTC time" [12:59] thats odd ok ill change it [12:59] k [13:00] ok saving [13:00] http://www.timeanddate.com/worldclock/fixedtime.html?month=6&day=22&year=2008&hour=18&min=0&sec=0&p1=0 [13:01] great :) [13:02] and sunbird updated already :) [13:06] Jazzva: ok i got your post, just one thing to add. If you add agenda item you should show up or it will be skipped and called at end of meeting if you still are not there i will be reviewed for another meeting time (god only know when that will be) i think i will add agenda item to regularize meetings maybe 1 a month or something [13:07] I'll be there :) [13:07] me too as of right this minute [13:07] i have my sister coming to town on the 15-16th and i think shes staying for a week so i should be good [13:08] Jazzva: i accepted your post to list since you are not a member [13:08] I'm not? [13:08] Jazzva: not from the email i got [13:08] Hmm... I thought I subscribed. [13:08] As list administrator, your authorization is requested for the [13:08] following mailing list posting: [13:08] List: Ubuntu-mozillateam@lists.ubuntu.com From: jazzva@gmail.com Subject: Mozilla Team meeting Reason: Too many recipients to the message [13:09] At your convenience, visit: [13:09] https://lists.ubuntu.com/mailman/admindb/ubuntu-mozillateam [13:09] maybe that email isnt but another one? [13:09] I don't use another one :)... [13:10] oh [13:10] dont know than [13:12] Ok, tried to subscribe again. It says I'm subscribed. Perhaps not everyone is allowed to post to the list :). [13:12] not sure you should have been [13:12] It's no prob... [13:12] i got post before the admin request [13:12] and than after it too [13:13] Jazzva: are you german or dutch? [13:15] Neither, I'm from Serbia :) [13:16] asac: is so maybe he will read his email and translate this bug [14:00] gnomefreak: hey [14:00] ff3rc1 is in intrepid, not RC2 which got released just couple of days ago. [14:00] AFAIK [14:03] oh yeah sorry, let me check something [14:04] ah thats why i have it [14:04] shirish: sorry i have RC2 in intrepid [14:04] shirish: it will be done sometime this week but as i recall not much changed from RC1>RC2 and not sure if anything was linux side [14:05] gnomefreak: you're subscribed to somebody's PPA, for I don't see it in the builds. [14:05] shirish: they will push it out as soon as they can, its a weekend though [14:05] shirish: i have my own as well i have like 4 PPAs i grab packages from [14:05] right. I understand that [14:06] gnomefreak: can u just mail me the changelog so I know what the changes that have been pushed through in RC2. [14:06] shirish: keep something in mind our PPAs always have latest snapshot and once we add our fixes they get uploaded to official repos [14:07] shirish: i can give you link i think [14:07] that would be cool enough [14:08] gnomefreak: I'm not looking to subscribe to the PPA, but definitely interested in knowing what changes to expect. [14:12] well mozillas release notes suck but here they are http://www.mozilla.com/en-US/firefox/3.0rc2/releasenotes/ as for our changes there hasnt been any yet from what i can tell just the .1 push [14:14] give me a minute and i will have our changelog [14:16] shirish: here is our changelog at the moment [14:16] shirish: http://pastebin.mozilla.org/455144 [14:16] this can be changed revised or totally forgotten about without any prior notice ;) [14:19] sweet [14:20] fta: did we remove DOM from our builds of Ffox-3 i know upstream did but i didnt see anything in changelog to match that [14:27] gnomefreak: thanx ;) [14:27] shirish: np [14:28] gnomefreak: there should have been some note about fsync too, from what I read some issues of fsync have been solved. [14:28] asac: fta looking at RC2 for fta's PPA we are still building DOM and upstream dropped it from source and is now a plugin should we do same? [14:29] shirish: it might have been fixed in another release i never saw that bug more than just a passing by [14:40] hi [14:40] the improvements are mostly in xul1.9 for us [14:41] yeah but i didnt feel like tracking that down too [14:41] * gnomefreak was hoping he knoew xul + ff == firefox [14:42] so much for RC2 being final [14:42] the fsync thing is in xul, not ff [14:43] 9 blockers atm [14:43] fta: that could be why i havent seen much on it [14:43] ? [14:43] the fsync [14:44] i was looking for it in ff not xul [14:51] fta: do you have the wiki for the improvments/goals for intrepid (for our pakcages) [14:51] https://wiki.ubuntu.com/MozillaTeam/Ideas/Intrepid [14:52] thanks [15:12] changed something ? [15:13] no [15:13] was looking for it the other day [15:14] i was hoping UDS stuff got on there [15:14] nope, no news from asac on that [15:15] ah ok [18:25] Jazzva, hm, since i've applied the asoundconf commands provided by crimsun and rebooted, i was happy with everything.. but teatime no longer has sound [18:26] What commands? [18:27] fta ^ [18:27] asoundconf reset-default-card && asoundconf set-pulseaudio [18:27] fta, that's in intrepid? [18:27] With some newer package? I'm still on hardy... [18:27] Can it be reproduced in hardy :)? [18:28] i've done that on intrepid [18:29] Ok. I'm doing an upgrade now. I think it contains the newer kernel version, so I'll have to restart. I'll test it in hardy after restart. [18:29] fta, did you see the messages I left in the morning (around 4, 5 am)? [18:31] yes but it's strange as my tests only show 0 and 2, not 2 and 3 [18:35] You printed the result of "gdk_window_get_state()" only and not "g_w_g_s() & mask"? [18:35] yes [18:36] Strange... [18:36] Did you print with fprintf(stderr, "%d", g_w_g_s(window_variable))? [18:37] I did it like that, and I got 0, 1, 2 and 3, depending on the sttate :/ [18:41] hm, i've rebuilt it and i see 0, 1, 2, 3 too [18:42] but that's with the patcj [18:42] h [18:42] The patch? [18:42] I tested with .14 ver... The current in hardy [18:42] - if((gdk_window_get_state(GTK_WIDGET(mainwindow)->window) & GDK_WINDOW_STATE_ICONIFIED) || !GTK_WIDGET_VISIBLE(mainwindow)) { [18:42] + if(!GTK_WIDGET_VISIBLE(mainwindow)) { [18:43] Oh... that patch :). I think I tried to change _ICONIFIED to _WITHDRAWN, and it didn't change the actual result [18:44] Then I went on looking in the gdk functions, looking if it's maybe a bug in g_w_g_s, but it looked fine. I also noticed they have enum mainwindowState, which uses values 0, 1, 2 and 3, but it's only used in ui_mainwindow_init(). So, I'm confused :). [18:47] window state shouldn't change before actual gdk_window_restore, or _hide (whatever is the name of that function). And it shouldn't return 0-3, but 1, 2, 4, ... :/. [18:49] no, it's a mask, so 3 = 2 + 1 [18:49] it sets bits [18:49] Oh... so, it can be in more than one state atm? [18:49] :) [18:50] yes [18:50] some states are not incompatible [18:50] Well, then that's good :)... [18:50] yes, i'm writing the patch [18:51] Yes, we don't need to change the condition in if, just add one more test in the if-branch [18:51] :) [19:05] hm, do we want 2 clicks or just one when the window is visible in another workspace ? [19:05] i'm used to 2 [19:05] Well, xchat uses that - 1 to hide window, and second to show it [19:06] Then we can just remove the first check, right? It worked that way last night [19:06] yep, that's what i'm used to too [19:09] dinner time. see you soon [19:18] ui [19:18] soccer fever ;) [19:19] hi [19:19] hi fta [19:19] i am heavily sunburned ;) [19:20] well not heavily, but quite OK :) [19:20] fta: on what no news? [19:20] people are looking for our blueprints [19:21] what except for largescalemaintenance do we need a blueprint for? [19:22] ok, but processing is done (upstream wise) [19:22] i have to redo the current ubuntu related docs about states and tags, but thats not a spec [19:22] ubufox is in flashexperience spec as well as firefoxkdeexperience [19:23] hmm ... maybe the addons thing could get a blueprint ... not sure though. [19:23] do we need a spec for daily builds?= [19:23] I'd say "just do it" [19:24] the graceful upgrade thing most likely wont happen. its not really understood. if someone comes up with bright ideas: let me know :) [19:24] i'm working on it, i needed the new mozclient for dyntag [19:24] (daily builds) [19:26] ok, i think we should link specs we already have on that page on top. [19:30] Jazzva, could you please test that: http://paste.ubuntu.com/18534/ [19:31] Jazzva, it's ok for me but it seems that function is sometimes called twice [19:32] asac, if you have time, i need help for flock, i'm out of ideas [19:32] fta: you sure you want to do _present? [19:33] afaik that doesnt really deiconify in all cases [19:33] though the behaviour might be what is wanted from the desktop side [19:33] well, it was called in the initial code, and it seems to do what it is supposed to [19:36] http://launchpadlibrarian.net/15023265/buildlog_ubuntu-intrepid-amd64.flock_2.0~a1~svn20080603r19205-0ubuntu1~fta4_FAILEDTOBUILD.txt.gz [19:36] http://launchpadlibrarian.net/15024875/buildlog_ubuntu-intrepid-i386.flock_2.0~a1~svn20080603r19205-0ubuntu1~fta4_FULLYBUILT.txt.gz [19:36] http://launchpadlibrarian.net/15025750/buildlog_ubuntu-hardy-lpia.flock_2.0~a1~svn20080603r19205-0ubuntu1~fta4~hardy_FAILEDTOBUILD.txt.gz [19:36] asac, ^^ [19:36] i've even tried shared clucene [19:36] https://code.edge.launchpad.net/~mozillateam/flock/flock.head [19:36] http://www.sofaraway.org/ubuntu/tarballs/flock_2.0~a1~svn20080607r19251+nobinonly.orig.tar.gz [19:37] it builds fine only on intrepid/i386+lpia, not intrepid/amd64 and hardy/* [19:39] /usr/bin/ld: /usr/lib/libclucene.a(FSDirectory.o): relocatio [19:39] yep, that's before the shared patch [19:40] ah [19:40] asac, here is with shared on hardy/amd64: http://paste.ubuntu.com/18038/ [19:41] fta: sure it picks everything from systme during build? [19:41] maybe change to that directory and just strace -eopen -f make [19:41] clucene is not included in the source [19:41] the file that fails [19:44] ok germany game starts ;) bbl [19:46] enjoy [19:46] http://jboriss.wordpress.com/2008/06/08/improving-tabs/ [19:49] fta@ix:~ $ liferea [19:49] dirname: extra operand `/usr/lib/xulrunner-1.9/libsqlite3.so.0' [19:49] Try `dirname --help' for more information. [19:49] lrwxrwxrwx 1 root root 13 2008-06-08 16:08 /usr/lib/xulrunner-1.9.1a1pre/libsqlite3.so.0 -> libsqlite3.so [19:49]