16:01:15 #startmeeting 16:01:15 Meeting started Thu Oct 17 16:01:15 2013 CET. The chair is API. Information about MeetBot at http://wiki.debian.org/MeetBot. 16:01:15 Useful Commands: #action #agreed #help #info #idea #link #topic. 16:01:28 #topic waiting 5 minutes just in case someone appears 16:06:18 well, 5 minutes 16:06:20 and a little more 16:06:25 so lets move to the first real topic 16:06:29 big turnout. 16:06:45 #topic "Boston" Summit Recap and Wayland situation 16:07:21 #info Last weekend (+Monday) some of the people of the a11y team were present at the "Boston" Summit, this year in Montreal 16:08:12 #info the main topic was Wayland, and we were being testing, debugging and triagging GNOME during the summit (and joanie even before) 16:08:48 #info and also talking with several developers there about Wayland situation. Specifically Matthias Clasen and Jasper 16:09:05 #info you can find a summary of current status and findings here: 16:09:12 https://wiki.gnome.org/Accessibility/Hackfests/Montreal2013 16:09:19 * clown looks 16:09:31 #info Joanie has started looking into the OSK too, which is not on that page yet. 16:09:36 #info we plan to use that page as a base of a more general and permanent "Current situaion of wayland" page 16:09:59 #info fwiw, we also fixed non-wayland bugs there. But as said, Wayland was the main topic. 16:10:02 done 16:10:21 There's also the mailing list threads. 16:10:21 so, any other want to add information about the summit? 16:10:34 ah true 16:10:38 I will info them 16:10:39 since the topic includes the Wayland situation 16:10:40 thanks 16:11:00 #info as an outcome of those conversations, some mailing threads were started 16:11:21 on the plus side, it looks like the magnifier is relying less and less on X. They got rid of XFixesCursor. 16:11:35 #info One in gnome-accessibility-devel about getting the information of the current window 16:11:37 https://mail.gnome.org/archives/gnome-accessibility-devel/2013-October/msg00006.html 16:12:03 #info and other on wayland list, sent by mclasen, that list several of the stuff we talked in the summit 16:12:09 https://mail.gnome.org/archives/gnome-accessibility-devel/2013-October/msg00008.html 16:12:34 clown, yes, afaik, that is part of the plan to get the magnifier working on wayland 16:12:44 there are other parts of gnome-shell that are being abstracted 16:13:06 although mclasen mentioned that jasper says that abstracting everything would be complex/impossible/not-worthy 16:13:18 and fwiw, I think that I'm done with the infos now 16:13:36 so again 16:13:40 so, any other want to add information about the summit? 16:13:52 Don't think so 16:14:02 other than the fact that we need to figure out if new a11y api is needed 16:14:08 and if so get that into place ASAP 16:14:16 and then get implementors to implement it 16:14:17 well, as I told you recently 16:14:22 is not only about new a11y api 16:14:25 but also about where 16:14:32 I know 16:14:35 I didn't went into all the details of last mclasen email 16:14:59 but if we don't prioritize this, things will be broken in 3.12 16:15:11 but he seems to suggest to add some api on wayland/mutter and not expose it through toolkits 16:15:23 fwiw 16:15:24 but that was before I responded with my other diea 16:15:26 idea too 16:15:30 taking into current state of wayland 16:15:42 yes, I don't really get what mclasen is proposing in his email. 16:15:53 I really would like having wayland on 3.12 as beta 16:16:04 not only for a11y reasons 16:16:23 having wayland "fully ready" for 3.12 sounds really too optimistic to me 16:16:29 fully ready==production ready 16:16:35 for me current status is not even beta 16:16:37 is alpha 16:16:49 I agree, but we cannot count on that happening 16:16:49 but I fear that what Im saying is just a rant 16:16:59 we have to assume that 3.12 is the target release 16:17:04 because that is what they have said 16:17:14 and that means figuring out this stuff NOW 16:17:19 and getting the API in place NOW 16:17:50 (and it's not a rant) 16:17:52 well, the good news 16:17:59 is that mclasen seems really interested 16:18:02 well, mclasen and others 16:18:08 and they are moving stuff 16:18:09 * joanie nods 16:18:26 so we are not alone here 16:18:36 so the action items? 16:19:30 ok, I can think on the following ones 16:20:12 #action API gnome-shell not accessible on wayland. Hacky solution avaible. Proper solution needed. 16:21:02 #action API: looking at tooltip related code to see the possibility of adding mouse-in, mouse-out event at the a11y toolkit 16:21:23 #action Joanie will keep testing the OSK and write up her findings, file bugs, etc. 16:21:27 #action everybody: being active on the threads opened 16:21:42 #action Joanie will create a non-hackfest/summit a11y in wayland page 16:21:58 mgorse: what are you working on w.r.t. wayland? 16:23:26 although probably off-topic on the current topic 16:23:36 we can't forget that other of the tasks for 3.12 16:23:44 and I removed that from the agenda 16:23:46 is about the async API for at-spi 16:23:47 * joanie quickly adds it 16:23:52 well, don't need 16:23:56 we can talk about that here 16:24:02 because is affecting the future planning 16:24:23 anyway, mgorse seems to be away so in any case: 16:24:30 * joanie puts "Other 3.12 work" on the agenda 16:24:38 and pretends it was always there 16:24:39 ;) 16:25:51 #info tentatively: mgorse will keep his focus on the at-spi2 async API, although he will properly reported by API and joanie about Wayland situation, discoveries and general stuff 16:26:05 sooo 16:26:10 anything else in this topic? 16:26:16 comments, questions, doubts? 16:26:21 fears? 16:26:23 :) 16:26:33 that's a lot to do by 3.12... 16:26:41 * joanie nods 16:26:47 * clown apologizes for his pessimism 16:27:02 technically is "just 2 topics" 16:27:08 my phrase for the whole of the summit was "Wayland zomg!" 16:27:10 the problem is when you start to list all subtopics ;) 16:27:13 with many many sub-topics. 16:27:41 in any case that is one of the reasons of my previous "gnome wayland should be beta for 3.12" rant 16:27:46 because, fwiw, 16:28:03 there are also a lot of subtopics there if we remove a11y from the equation 16:28:04 I don't know if it matters, but there is a "window manager" insides gnome-shell's "shell-global.c", that makes exttensive use of the the XDisplay... 16:28:48 clown, well, one of the things that makes the move to wayland complex 16:28:57 probably much too specific for this meeting. 16:29:05 is that they would need to support apps using still X through xwayland 16:29:17 eww. 16:29:20 that means that wayland environment will not be "pure wayland" 16:29:30 in any case, probably that would not be the case for gnome-shell 16:29:44 and that X stuff will be abstracted, ifdefed or something 16:30:38 so after this rant-alike chat about wayland 16:30:40 moving to next topic? 16:30:49 sure 16:31:13 #topic W3C updates 16:31:19 clown? 16:31:36 ah wait 16:31:44 * API refreshing agenda 16:31:56 #topic Other 3.12 work 16:31:58 ok 16:32:04 sorry clown, 16:32:21 #info plans for 3.12 are similar to the list we made on last week 16:32:32 #info Main topics: wayland + at-spi2 async API 16:33:03 and possibly at-spi2/X/wayland interactions. 16:33:04 * joanie wonders about the focus deprecation stuff 16:33:05 ? 16:33:25 #info Medium topics: already existing features (tinting+focus tracking UI) 16:33:38 focus is deprecated? What does that mean? 16:33:44 #info Medium topics: that to-be-proposed feature about homogeneous keyboard navigation 16:33:52 ah yes, lots of activity on the tinting bugzilla. 16:34:01 clown, well, I include at-spi2/X/wayland interactoin on wayland 16:34:09 okay. API 16:34:14 clown, well, not focus 16:34:22 is what we said last week 16:34:28 we had AtkObject:focus-event 16:34:35 and AtkObject:state-changed:focus 16:34:36 event 16:34:39 we deprecate one 16:34:40 i was not here last week... 16:34:46 * clown reads. 16:35:01 this is something we want to tackle early on in the cycle 16:35:02 and we also deprecated several utility methods related with focus tracking 16:35:05 oh, right, I remember that from two weeks ago. 16:35:10 in order to deal with regressions 16:35:16 in any case, 16:35:18 and I don't want us to lose site of that 16:35:20 * API resuming infos 16:35:23 during the Wayland zomg 16:35:31 they are both ATK focus events, right? 16:36:00 and somehow parallel "desktop focus" events. 16:36:04 #info Medium topics: several focus related methods were deprecated. Some work started on toolkits: test it 16:37:14 #info Medium topics: atk_text_get_string_at_offset defined at ATK. Need to start to implement and test it on the toolkits. Probably that would lead to having just one implementation on gtk/clutter, probably on pango. Follow up this, test it 16:37:28 #info about the previous: webkitgtk already has a implementation of that method 16:37:40 phone 16:37:48 * clown waits 16:38:13 Regarding pango, I thought the long term goal was to no longer depend on pango. 16:39:16 sorry, back 16:39:29 joanie, well, that is the reason I added "probably" 16:39:41 I know that the plan was removing totally pango from webkitgtk 16:39:55 but I still don't know what is the real plan for gtk/clutter 16:40:02 depends on the week I hear one or the other 16:40:05 in any case 16:40:19 that was my mental list of main and medium topic for 3.12 16:40:28 did I miss something? 16:40:40 action items? 16:41:02 well, I think that I already have an action item about the to-be-proposed feature 16:41:10 and pinging benjamin 16:41:21 and I already have some action items about some subtopics of wayland 16:41:24 I'll take one for testing his patch in process 16:41:33 k 16:41:49 #action Joanie will test Benjamin's work-in-progress focus-related deprecation patch 16:41:52 #action API: check with gtk and clutter developers what are the plan towards pango 16:44:06 in any case 16:44:21 remember that we are at the beginning of the cycle 16:44:25 and listing all the stuff to do 16:44:35 I have just classified on Main and Medium topics 16:44:42 and panicking about it 16:44:43 ;) 16:44:49 but probably we would need to eventually do a more fine-grained prioritization 16:44:56 and see what can wait to 3.14 16:45:00 so we can panic in order 16:45:39 I will info it just in case 16:45:56 #info we are at the beginning of the cycle, so listing all the remaining tasks 16:46:15 looks comprehensive to me API. 16:46:22 #info eventually we would need to do a fine-grained prioritization and see if there is something can way to 3.14 16:46:26 and having said so 16:46:33 is a good moment to change topic 16:46:44 #topic W3C updates 16:46:49 clown 16:46:57 ? 16:47:01 #info The UAIG document is very close to going to last call status. 16:47:17 #info there are a few more edits to be made, and about 20 or so tests to run. 16:47:35 #info the plan is to have everything wrapped up by the end of Oct 16:47:54 on a related topic... 16:48:02 #info Joseph gave a talk at the jQuery accessiblity summit last week about ARIA. 16:48:11 #info http://www.deque.com/deque-partners-jquery-create-accessibility-summit 16:48:20 yay! 16:48:23 how'd it go? 16:48:46 it went well, although I had a cold the weekend before, and lost my voice for about the last two minutes of talk. 16:48:51 turned into a croak. 16:49:00 oh dear. 16:49:06 But 'grats on the "went well" part 16:49:07 :) 16:49:38 otherwise, it went okay, however, the crowd was mostly developers, and I got a feeling of: we want to do things that work, and specs just get in the way. 16:50:00 clown, so most of the questions they made where in fact complains? 16:50:18 although some of the webaim people reinforced some of what I was saying. 16:50:37 API, the main negative aspect were comments from Hans Hillen. 16:51:14 He stated that he is only concerned with screen readers, and find aria lacking at times both the spec and browsers implementation of it. 16:51:25 my reply was: I find JAWS lacking in terms of its support of ARIA. 16:51:36 heh 16:51:45 Orca's ARIA support is on my to-do list 16:51:53 it exists but needs improvement 16:51:54 so I guess that Hans Hillen is a JAWS developer? 16:52:18 Hans is TPG (The Paciello Group) person. 16:52:38 He does a lot of jQuery coding, and consutling on websites about their overall accessibility. 16:52:58 It's good in the sense that you can see what works and doesn't work in "real world situtaions". 16:53:02 ah well, I asked that due your answer 16:53:11 I thought that was a kind of "burn" answer 16:54:03 kind of: I checked some of his complaints and FF (at least) was publishing the information properly to AT-SPI. I assume it does as well on Windows. 16:54:06 anyway, saying publicly "I'm only concerned with screen readers" sounds somewhat odd to me 16:54:11 So, how it JAWS missing It? 16:54:33 he was proposing that aria should just focus on it? 16:54:43 almost, API. 16:54:44 s/aria/w3c 16:55:18 More so aria, and the WAI (Web A11y Iniiative). 16:55:26 the WAI is a group within the w3c. 16:55:53 I invited him to post to the W3C mailing lists about any ideas he might have. 16:56:04 we'll see. 16:56:33 Before I forget, what is your status vis-a-vis W3C joanie? 16:56:53 Igalia is going to apply to join (like this week I hope) 16:57:04 and in the meantime I will be granted a six month IE status 16:57:23 because on Tue there was the "emergency" meeting called for tomorrow that you might be interested in. 16:57:24 I've pinged folks internally to start the ball rolling on the application 16:57:43 "emergency"? 16:57:56 Yeah, I endorse that question mark. 16:58:11 heh 16:58:30 The "problem" is that various people in the W3C are proposing implemenation details for a11y, including ARIA. 16:58:32 without any prep time, along with the status not yet in place, I'm not sure if I should go 16:58:42 But, they are doing it independently with one another. 16:58:50 great 16:59:07 So, RIch is worried this might get out of hand, and called this meeting to get all of us on the same page. 16:59:08 implementations details? 16:59:10 and if it doesn't get solved *right now* then it's in stone? 16:59:17 ah 16:59:21 The title of the meeting is: "Accessibility Implementation Guide coordination" 16:59:28 Getting people on the same page doesn't require me. 16:59:48 Sure. I just wanted to let you know in case you were interested. 16:59:49 hmm, I don't see how a spec that can be implemented on so different systems 16:59:54 can have a implementation guide 17:00:05 clown: thanks! 17:00:07 unless I'm thinking on "implementation guide" as something too low level 17:00:10 ? That's what the UAIG is API. 17:00:19 I think I'll pass on this one, but start getting involved soon 17:00:21 The "user agent implementation guide". 17:00:45 ah ok 17:00:53 then yes, is "high level" enough to me 17:01:04 It has tables for how aria is to be implemented on each a11y platform. 17:01:24 clown, ok 17:01:36 sorry, probably Im somewhat mental-vague today 17:01:49 I have just came back to spain yesterday ;) 17:01:51 API needs more sleep 17:01:51 problem is there are people in the HTML5 working group who are doing similar things, and are sometimes inconsistent with the ARIA guide. 17:02:02 yes, he does. 17:02:47 anyhow, that's all from me about w3c. 17:03:00 ok 17:03:10 so, joanie more questions? 17:03:15 nope 17:03:19 ok 17:03:23 so clown, thanks for the update 17:03:30 I will skip marketing because jjmaring is not here 17:03:33 wlcm 17:03:34 *jjmarin 17:03:46 #topic "Miscellaneous Time" 17:03:53 anything not scheduled to mention? 17:05:17 crickets! 17:05:21 crickets everywhere! 17:05:29 * clown *chirp* 17:05:31 so having said so, I will end the meeting 17:05:34 #endmeeting