15:09:34 #startmeeting 15:09:34 Meeting started Thu Feb 2 15:09:34 2012 UTC. The chair is API. Information about MeetBot at http://wiki.debian.org/MeetBot. 15:09:34 Useful Commands: #action #agreed #help #info #idea #link #topic. 15:09:50 #topic Hackfest Summary 15:10:16 #info API made two posts with a summary of the hackfest 15:10:21 http://blogs.igalia.com/apinheiro/2012/01/19/atkat-spi2-hackfest-2012-day-1/ 15:10:26 http://blogs.igalia.com/apinheiro/2012/01/24/atkat-spi2-hackfest-2012-days-2345/ 15:10:48 #info but being really short 15:11:00 #info this time was more balanced between discussion and hacking 15:11:06 #info although still more discussion 15:11:31 #info still a lot of questions, mostly from Benjamin side, as he wants to change a lot of things, but not sure about the way to do that 15:11:46 #info one of the main conclusions is going to have accessibility enabled by default 15:11:56 #info more testing, from the early stages, etc 15:12:14 #info and the first step of that plan, stop to use atk-bridge as a module/plugin 15:12:26 #info still not clear specific implementation: 15:12:41 #info a library? => new dependencies to gtk, cluttter etc 15:13:01 #info introduce it into atk => atk will be bigger, will have a dependency with at-spi2-core 15:13:19 #info gtk talking directly on the BUS => only benjamin likes that idea 15:13:55 #info the idea would be talk about that now, and start it at the beginning of 3.6 cycle 15:13:59 done 15:14:06 other wants to share any summary of the hackfesT? 15:15:35 gtk talking directly on the BUS means removing atk ? 15:16:32 jjmarin, imho, just for gtk 15:16:48 I don't see any reason to do the same for clutter, for example 15:16:51 I guess gtk wouldn't use it at least; it's an open question as to what FF/LibreOffice/Clutter would do 15:16:58 taking into account current clutter requirements 15:18:43 what does talking directly on the BUS mean? how does it by-pass atk? isn't atk just a set of interfaces? 15:19:20 clown, right now atk-bridge 15:19:27 uses atk 15:19:28 clown: It's a set of interfaces, but we have atk-bridge that exposes a11y over d-bus via atk 15:19:33 atk objects and signals 15:19:59 clown: yes, atk is just a set of interfaces - but it's not the interfaces that are used by at-spi, just a rough match 15:20:03 so, we aren't talking about removing atk, but removing or replacing atk-bridge. 15:20:29 or replacing atk interfaces with at-spi interfaces (?) 15:20:31 clown, we aren't talking about removing or replacing atk-bridge 15:20:34 clown: and the question remains: what's the purpose of ATK when what really matters is the dbus interface 15:20:44 just that this is and option still in consideration for gtk 15:20:55 clown: also, qt doesn't use atk either 15:21:14 Company, but they have their own accessibility interfaces 15:21:25 not ATK, but sill, accessibility interfaces 15:21:39 yes 15:21:56 'one interface to rule them all, one interface to bind them' :-) 15:22:10 * joanie laughs 15:22:16 but the question is: why have an intermediate step with ATK? 15:22:25 just to make mordor jokes? :p 15:22:29 :-) 15:22:54 Company, to avoid having a gtk-bridge, a clutter-bridge, a firefox-bridge, etc. 15:23:08 the point behind the mordor joke was "we" need an interface. There's ATK, there's AT-SPI, there's IA2, etc... 15:23:13 replace bridge for dbus calls if you want 15:23:20 but we have that 15:23:21 one common interface would be nice. 15:23:28 we currently have a bridge to atk everywhere 15:23:38 it's called gail 15:23:42 or cally 15:23:56 i thought you said qt doesn't use it. 15:24:10 no, qt bridges to dbus directly 15:24:36 but, if I want to call something to say, get the accessible name in a qt environment, what do I call? 15:24:38 Company, as far as I understood what I asked to frederik on the hackfest 15:24:47 they also have that accessibility interfaces 15:25:05 yes 15:25:13 in that sense, they could write a qt-bridge based on CORBA and reusing they accessibility interface implementation 15:25:18 as GNOME did in the past 15:25:25 there is a qt a11y abstraction 15:25:37 interestingly, both webkit and mozilla have that, too 15:26:04 and while the browsers chose to bridge that to atk, fregl chose to bridge qt's interfaces to dbus directly 15:26:36 hi 15:27:01 hm, I keep on losing my nick 15:27:20 Company, because after all, bridging to atk would mean that you wouldn't need to worry about DBUS, as this was managed by atk-bridge 15:27:26 is like webkitgtk folks decided 15:27:39 yes, you have to worry about atk instead 15:27:40 that it would be easier to bridge to atk than write a new bridge 15:27:54 API: are you worried that DBUS may someday go away? 15:28:11 decisions decisions, life is full of decisions ;) 15:28:12 that would be a reason to have some intermediate layer. 15:28:19 clown, not really, that would be really unlikely 15:28:29 but yes is a reason to maintain the bridge 15:28:47 but it is not the only reason, IMHO 15:28:59 ATK has all the benefits and drawbacks that abstraction layers tend to have 15:29:07 I still think that it is easier to implement ATK 15:29:10 Company, exactly 15:29:18 they shield you from the pain of the underlying system 15:29:25 but also from the benefits 15:29:27 I know that choosing one or the other can be tricky 15:29:42 but I still think that at this moment it would be easier and more efficient 15:29:49 to improve ATK and the bridge 15:30:00 to get what we lack right now 15:30:13 instead of do a new bridge 15:30:22 but as Im saying, just my opinion 15:30:57 anyway, I'm the evil entity in chair of this meeting 15:31:05 I think that this is not the place to go so deep 15:31:14 so unless someone want to add something else 15:31:19 I will go on with the meeting 15:31:19 yeah, #a11y is open for these questions all week :) 15:31:41 I will take that as a "go on" 15:31:50 #topic 3.4 features update 15:32:01 clown, last week we didn't have meeting 15:32:12 so you can provide a summary of your features now 15:32:18 so I didn't miss it afternall? 15:32:25 "after all" 15:32:26 no, lack of people 15:32:29 sure, API. 15:32:47 this is in regards to 3.4 features... 15:32:55 #info Zoom Options Dialog, launched from control centre's Universal Access Panel 15:33:01 #info https://live.gnome.org/ThreePointThree/Features/ZoomOptionsDialog 15:33:09 #info Merged into master as of 21 Jan 2012. 15:33:15 #info https://bugzilla.gnome.org/show_bug.cgi?id=643086#c46 15:33:15 04Bug 643086: normal, Normal, ---, control-center-maint, RESOLVED FIXED, Universal access: add dialog for zoom options 15:33:22 #info note: one minor change made by Bastien before committing: he made the dialog modal. 15:33:28 #info I'm not sure about that, but it's not enough of an issue to argue. 15:33:34 #info Probably to be consistent with other control centre dialogs. 15:33:52 #info Lightness inversion, brightness, and contrast effects 15:34:01 #info https://live.gnome.org/ThreePointThree/Features/LightnessBrightnessContrastEffects 15:34:26 #info Settled on new algorithm for brightness/contrast -- the one that GIMP uses (as earlier suggested by Neil). 15:34:34 #info Have patched clutter-1.8 with the new code, and uploaded to bug (https://bugzilla.gnome.org/show_bug.cgi?id=656156#c26) 15:34:34 04Bug 656156: enhancement, Normal, ---, clutter-maint, UNCONFIRMED, Add lightness, brightness, and contrast effects 15:34:52 #info clutter-1.8 branch because that is what GNOME Shell jhbuild does, according to its wiki docs. 15:35:03 #info Can compile and link with clutter master. 15:35:24 #info Still need to test master branch version. 15:35:31 #info but, ... ebassi has suggested using the new CoglSnippet API to implement all three effects. 15:35:37 #info see https://bugzilla.gnome.org/show_bug.cgi?id=656156#c24 15:35:37 04Bug 656156: enhancement, Normal, ---, clutter-maint, UNCONFIRMED, Add lightness, brightness, and contrast effects 15:35:43 #info I will investigate this technique. 15:36:02 #info and, I hope to have a new patch for master branch by tomorrow. 15:36:07 #info Focus/caret tracking 15:36:14 #info https://live.gnome.org/ThreePointThree/Features/FocusCaretTracking 15:36:20 #info no progress due partly to other obligations (the above, my aria work). 15:36:31 #info also, the IDRC (my employer) is reluctant to proceed without funding. 15:36:43 #info I have suggested to exchange another magnifier feature for the focus work. 15:36:51 #info for example, one of the features was to support multiple monitors. 15:37:02 #info so: consider delaying multiple monitor support until after focus tracking (possibly need to clear this with AEGIS). 15:37:26 #info will likely proceed to this work after finishing up the inversion, brightness, and contrast effects. 15:37:31 * clown done (whew!) 15:37:37 whew :P 15:37:42 probably the longest summary 15:37:51 clown, do you plan to update the wiki, right? 15:38:01 stretching the definition of "summary"... 15:38:11 the 3.4 feature list, API? 15:38:15 yes 15:38:20 I have made some mods since last week. 15:38:38 I will update the colour effects after I get a new patch attached. 15:38:43 clown, ok, thanks 15:38:48 no problem., API 15:38:51 so, joanie ? 15:38:54 about dots? 15:39:17 #info Joanie did much of the work on the Dots introspection (still in her sandbox). BUT has had to focus on Orca. 15:39:34 #info Joanie really and truly hopes to finish the introspection work soon and get a patch to Fer. 15:39:36 (done) 15:39:44 (and sorry for getting tied up with Orca) 15:40:07 ok, anyway, it would be good if you also update the wiki 15:40:22 clown, joanie and as it is about 3.4 features 15:40:40 it would be good to say something like "it will/it will not be ready for 3.4" or something 15:41:38 so as no other one here has any 3.4 feature in his plate I will move to next topic 15:41:40 gotcah 15:41:47 * clown can't type. 15:41:47 #topic Q4's are due 15:42:25 #info Emily Chen sent the usual "GNOME teams needs to write what they did on Q4" 15:42:25 #info Your notes go here: https://live.gnome.org/Accessibility/QuarterlyReports/2011/Q4 15:42:37 perfect timing ;) 15:42:40 #info Piñeiro and I will write up and submit the final report. 15:42:43 (done) 15:42:44 hehe 15:42:46 I think that this is easy 15:42:50 any questions or doubts? 15:43:10 any link to the last one Q3? 15:43:27 https://live.gnome.org/Accessibility/QuarterlyReports/2011/Q3 15:43:30 and.... 15:43:32 right. 15:43:41 https://live.gnome.org/GnomeMarketing/QuarterlyReports/2011/Q3 15:43:50 basically, it's always the same 15:43:56 just change the numbers 15:44:48 ok, no questions 15:44:49 lets move 15:44:53 #topic Marketing and Fundraising 15:44:55 jjmarin, ? 15:45:20 hi ! 15:46:09 #info GNOME was submitted to the The Computerworld Honors Program with special remark about accessibility 15:46:51 #info the Friend of GNOME progress bar has been put in www.gnome.org 15:47:30 #info By now we have US$ 4263 15:48:32 #info A piece of news about the a11y hackfest is in the GNOME news queue 15:48:40 and I think is all 15:48:59 ok thanks 15:49:17 anyone has some questions or doubts about marketing? 15:49:48 I want to push to put the progress bar in pgo and friend of gnome, but with many people in FOSDEM will be difficult 15:51:44 ok, thanks 15:51:49 I think we can move on :-) 15:51:54 taking into account that there are no more questions 15:51:55 yes 15:52:00 #topic miscellaneous time 15:52:05 I will start 15:52:16 #info API will go this weekend to FOSDEM 15:52:39 #info he will have a talk about accessibility 15:52:41 done 15:56:28 * jjmarin realises that keyboard navigation of gnome-shell is not quite good 15:56:49 do you know if there is any action to improve this ? 15:57:10 to improve what specifically? 15:57:17 why do you say that it is not quite good? 15:57:31 in my case, I don't like too much 15:57:44 need to press up and down to move to the left and right on the overview 15:57:57 but it is just about taste 15:58:12 did you find something more else¿ 15:58:46 can one use accerciser to test keyboard a11y support? 15:59:19 if so, could one use it to evaluate gnome-shell? 15:59:31 clown, well, you can start to use keyboard, and see if accerciser is receiving any event 15:59:40 but again, jjmarin 15:59:54 why do you say that gnome-shell key nab is not good¿ 15:59:56 ? 16:00:14 For example I don't understand why I can't use they key when I go to the apps view and start using the keyboard directly. I have to access the app using the accessibilty switch to start using the keyboard 16:00:55 jjmarin, because default view is search vie 16:00:59 if you press alt+f1 16:01:08 the one that receive the focus is the search entry 16:01:33 so this is not a problem about key nav 16:01:46 it is just that seems that you would prefer a different default view 16:02:01 if you do alt+f1 16:02:01 joanie, great! If you need some help with it or finishing I can spend some time with it! 16:02:07 entry gets the focus 16:02:09 (sorry for the delay) 16:02:12 so you can enter text 16:02:20 if you enter text a search starts 16:02:29 and you can use up+down to move on the search 16:02:54 in the same way 16:03:01 a mouse user 16:03:12 needs to use the mouse to click on applications to go to that view 16:03:23 I think If I go to the apps view if because I want to use the categories. I can use the search from the windows view 16:03:28 Okay; I need to leave. Back later. 16:03:52 jjmarin, but as I said 16:03:58 if a mouse user want to do that 16:04:05 he also needs to use several clicks 16:04:17 key nav allow to do that 16:04:21 so again, why it is wrong? 16:04:51 anyway, we are out of time 16:04:58 and this discussion is too disperse for that 16:05:03 so resuming at #a11y 16:05:05 yes 16:05:06 and closing the meeting 16:05:10 #endmeeting