15:59:37 #startmeeting 15:59:37 Meeting started Thu Jul 26 15:59:37 2012 CET. The chair is jhernandez. Information about MeetBot at http://wiki.debian.org/MeetBot. 15:59:37 Useful Commands: #action #agreed #help #info #idea #link #topic. 15:59:58 #topic GNOME 3.6 16:00:31 okay, jhernandez, here goes. 16:00:39 clown: ACK 16:00:41 :] 16:01:05 #info Joseph fixed the main caret tracking problem in his focus tracker, based on mgorse's suggestions. 16:01:17 #info the code now tracks the caret. 16:01:58 #info Joseph, put together a patch, uploaded it to the bugzilla, and emailed the a11y list. 16:03:07 #info there is still an issue of a delay when focussing on a gnome shell widget (St), and when the Atspi event is returned, it contains no accessible (? or maybe no component). 16:03:24 #info next step is to try and write it in C. 16:03:42 * clown looks up bugzilla to see what the issue is again with gnome-shell widgets. 16:04:32 #info there is still an issue of a delay when focussing on a gnome shell widget (St), and when the Atspi event is returned, it contains no component. 16:04:49 * jhernandez wondering about why clown needs to move this code into C 16:05:08 because it might work? 16:05:10 :-) 16:05:26 clown: ACK 16:05:53 take a look at what API said: https://bugzilla.gnome.org/show_bug.cgi?id=647074#c6 16:05:53 04Bug 647074: enhancement, Normal, ---, gnome-shell-maint, UNCONFIRMED, GNOME Shell Magnifier should track focus and the caret 16:06:05 thanks 16:06:08 * jhernandez reading ... 16:06:21 he doesn't say that it will wokr in C, but that it may be a factor. 16:07:05 on the other hand, if I rewrite in C and get exactly the same problem, then there is a bug somewhere. Either in atspi or in St, or in the communication between the two. 16:07:34 * jhernandez got it 16:07:39 so jhernandez, do you think it should work in JavaScript? 16:07:50 do you see any reason for the problem? 16:08:02 off hand, that is. 16:09:04 I thought that gobject introspection stuff was working fine with javascript 16:09:05 I really should try joanie's stand alone python version, first. 16:09:20 yeah, so did I. 16:09:21 but, I'm a little more pythonic developer 16:10:00 well, if you see anything, let me know. 16:10:28 shall I go on? I have updates for zoom options in 3.6 16:10:44 clown: I think joanie's suggestion was the best here 16:10:55 s/was/is 16:11:07 what is her suggestion? 16:11:18 < clown> I really should try joanie's stand alone python version, first. 16:11:55 ah, yes. I see. (it's still morning here). 16:12:10 * clown takes his second sip of coffee. 16:12:15 hehe 16:12:20 no prob 16:12:23 anyhow, moving on? 16:12:28 clown: of course 16:12:43 :] 16:12:50 #action Joseph to confirm that joanie's python standalone focus tracker demo still works. 16:13:35 #info Joseph has uploaded two glade files showing possible UI's for the new color effects in the magnifier. 16:13:39 * clown looks up bug... 16:14:36 #info, FWIW, Joseph favours this one: https://bugzilla.gnome.org/show_bug.cgi?id=676817#c5 16:14:36 04Bug 676817: normal, Normal, ---, control-center-maint, UNCONFIRMED, universal access: Allow to control color-related shader options in the shell 16:14:52 #info a screen shot of the above is here: https://bugzilla.gnome.org/attachment.cgi?id=219602&action=edit 16:15:26 #info Joseph has been working on modifying the gnome-control-panel zoom-options.c code to make use of the porposed UI. 16:16:01 #info that code is nearly ready; hopefully there will be a full patch to attach to bug soon. 16:16:11 jhernandez: do you use Fedora? 16:16:26 nope, I currently use openSUSE 16:16:37 why? 16:16:46 maybe you know the answer. 16:17:26 maybe not 16:17:29 when I run the latest nightly of gnome-shell, gnome-control-center, and gsettings-desktop-schemas, I get a bug out of dconf. 16:17:32 :-) 16:17:33 or maybe for testing purposes? 16:17:48 * clown looks up error message. 16:18:26 "** (process:17779) : WARNING **: failed to commit changes to dconf: GDBus.Error:org.freedesktop.DBus.Error.UnknownMethod: No such method 'Change'. 16:18:35 I have my hackbox here, so I could give a try to this 16:19:01 API suggested that if I launched from the 'recovery terminal' I could get past this. 16:19:40 unfortunately, he uses ubuntu, which has an recovery terminal option from the login screen. There he does: 16:19:52 jhbuild run dbus-launch --exit-with-session gnome-session 16:20:06 yup 16:20:25 Fedora's login screen doesn't offer that option. So, I'm puzzled as to how to make do the dbus-launch. 16:20:42 I can do it from any old terminal *after* I log in, but that doesn't help. 16:21:10 it's "too late", I think. 16:21:32 hmm 16:21:50 aren't you glad you contacted me for a light meeting? 16:21:53 '-) 16:23:34 of course! 16:23:54 sorry, I was moving to another room 16:24:13 * jhernandez looking for a link 16:25:00 * clown hoping you find it 16:25:25 ey 16:26:36 well, the link I was looking for is about how to get running your gnome-session under jhbuild from scratch 16:27:22 ah, okay. 16:28:34 about the dconf error message ... maybe out-of-date dconf version? 16:29:05 dunno 16:29:27 or, maybe asking the wrong dbus? 16:29:42 actually, dconf is "too new". 16:30:03 rather, the system's dconf is "too old" 16:30:49 when I do 'jhbuild run gnome-shell —replace", I'm running the gnome-shell under its "install" directory 16:31:14 and that environment expects to use the latest dconf, but that's not the one running. 16:31:28 API's technique is a way to get the new one running from log in. 16:31:55 * clown looks up API's email. 16:32:20 quoting API: 16:32:22 In fact I think that the problem is due last 16:32:22 changes on DConf, so if you are using gnome-shell and all stuff from 16:32:22 jhbuild, it doesn't work fine with the settings-daemon from the system. 16:32:22 So in summary you should be using gnome-settings-daemon from jhbuild 16:32:23 too. In order to get that I login using the recovery terminal and then I 16:32:25 execute a new session using my jhbuild stuff: 16:32:27 jhbuild run dbus-launch --exit-with-session gnome-session 16:34:11 well, I used to run the latest gnome code by adding a new session to the system and then logging into the session by selecting my jhbuild's gnome session as default session 16:34:29 too "sessions" i think ... 16:34:32 :P 16:34:48 * jhernandez looking again for this link 16:35:11 clown: how is going the implementatino of tracking? 16:35:27 I'm thinkin in implement a skel of a "emitter" of tracking information 16:35:28 into gs 16:35:36 in C 16:35:38 GObject 16:35:40 glib 16:35:49 aleiva, not much news — I was going to try and port the stuff to C/GObject, and see if that worked better. 16:35:50 http://developer.gnome.org/jhbuild/stable/jhbuild-and-gnome.html.en 16:36:18 But, my time is currently taken up by the prefs dialog for the new brightnesss/contrast effects in gnome-control-center. 16:36:18 clown: I did it by following these instructions 16:36:21 * clown reads log. 16:36:56 I think we are thinking the same thing, aleiva (re: focus tracking C/GObject). 16:36:58 clown: what do you think if I try that "emitter" thing and you imelement the client 16:37:05 and settings 16:37:13 and gs stuff that I think you know better than me 16:37:18 after the contrast bug :-P 16:38:06 aleiva, I kind of wanted to try it, but, in the interest of time, go ahead. 16:38:49 * clown takes a look at jhernandez's link 16:39:59 aleiva, you are going to make the focus/caret tracker such that it will support connect()? 16:40:19 something like that 16:40:25 or callback like 16:40:35 whatever gnomeshell people usually do 16:40:42 in this cases 16:40:45 that's what I was trying to do with the JavaScript version — receive events from Atspi, and relay them out using the standard gnome signal technique. 16:41:36 so, my focus tracker emit()s events, and allows clients to connect()/disconnect() to the tracker. 16:41:51 yeah, that's the logical way 16:42:15 but, on the atspi side, it uses the register() technqiue. 16:42:53 okay. give me a footnote :-) 16:44:33 I'm at guadec and I can talk with gs people I think 16:44:37 that will boost this 16:44:40 I hope 16:44:51 I only have one free weekend :-P 16:45:01 aleiva: go for it! 16:45:32 indeed. 16:48:43 jhernandez: re: the jhbuild instructions, is the relevant section 3.3? 16:50:32 clown: yes 16:51:13 maybe this could help you 16:51:14 dunno 16:51:33 maybe, I'll see what I can do. 16:51:57 ok 16:51:58 3.3.1 looks like the important part. 16:52:52 * jhernandez nods 16:53:40 and 3.3.3 16:54:02 * clown looks up his .jhbuildrc file 16:54:49 okay, I have moduleset = 'gnome-apps-3.6' in my .jhbuildrc. 16:55:48 I'm thinking I may not need to create this "gnomedev" account, but just to massage my own account as they suggest (and hope that doesn't bork the machine). 16:56:27 clown: maybe yes ... maybe not :P 16:56:48 I always create another account for this kind of purposes 16:57:00 I usually bork my machine 16:57:07 :P 16:57:32 yes, another account is safer. But I want to avoid having to re-build all of 3.6. It takes hours. 16:57:42 yes 16:58:19 anyhow, on another topic, I think the meeting is over? 16:58:43 * jhernandez nods 16:58:55 actually, let me record some of the things that aleiva just said. 16:59:42 #info regarding focus tracking, aleiva is going to try to write the tracker in C/GObject/glib 17:00:10 #info he is going to follow the architecture in Joseph's JavaScript tracker: 17:00:50 #info taking Atspi events, and exposing them using the standard gnome signal technique (emit(), connect(), disconnect(), etc). 17:01:07 done, unless aleiva wants to add anything. 17:01:22 nothing more 17:01:25 its ok, thanks 17:01:36 ok magnifier guys! 17:01:55 okay, thanks for the collaboration, aleiva! 17:01:57 :] 17:02:16 so ... 17:02:29 clown: do we have more topics?¿ 17:02:35 not from me. 17:02:46 * clown double checks 17:03:16 #topic Marketing and Fundraising 17:03:33 :-D 17:03:40 * clown wonders where jjmarin is. 17:03:44 #info today we reached the FoG goal! 17:03:46 :] 17:04:00 xD 17:04:16 jjmarin was in a webkit related talk 17:04:18 cool! 17:04:26 so .... 17:04:46 #topic Miscellaneous time 17:04:57 \o/ 17:05:01 * aleiva steal the gavel 17:05:03 clown: do you have some pretty miscellaneous topic? 17:05:15 nope. 17:05:25 clown: perfect 17:05:33 we (aleiva and me need a coffeee) 17:05:34 :] 17:05:36 ~o/ 17:05:50 clown, and the misc joke? 17:05:58 have a good one, guys. 17:05:58 xDDD 17:06:00 where's it? 17:06:06 why did the chicken cross the road? 17:06:15 no idea 17:06:24 because it was too far to walk around. 17:06:29 xDDDD 17:06:30 xDDD 17:07:36 ok, guys, let's end the meeting 17:07:38 :] 17:07:39 3 17:07:41 2 17:07:43 . 17:07:45 . 17:07:48 . 17:07:50 1 17:07:53 #endmeeting