16:06:12 <API> #startmeeting
16:06:12 <tota11y> Meeting started Thu Nov  8 16:06:12 2012 CET.  The chair is API. Information about MeetBot at http://wiki.debian.org/MeetBot.
16:06:12 <tota11y> Useful Commands: #action #agreed #help #info #idea #link #topic.
16:06:32 <API> #topic FoG update
16:06:36 <API> joanie, ?
16:06:41 <joanie> hey
16:06:44 * joanie infos
16:07:00 <joanie> #info Joanie pinged Bastien about the proposals but got no reply.
16:07:22 <joanie> #info Joanie pinged Karen to see if she would resume this task and she said she would share it with Bastien.
16:07:32 <joanie> #info Joanie forwarded the original proposals to the Board list.
16:08:08 <joanie> #info Karen initially responded, stating that we'd discuss these proposals on the Board list and/or the upcoming meeting.
16:08:17 <joanie> so slooooooooow progress
16:08:18 <joanie> questions?
16:09:01 <jjmarin> maybe next week :-)
16:09:15 <joanie> next week?
16:09:41 <jjmarin> I guess there will be more infor for questions
16:09:54 <joanie> perhaps; perhaps not
16:10:01 <joanie> we meet every two weeks
16:10:09 <joanie> due to Sandy we got off by a week
16:10:13 <joanie> our last meeting was Tuesday
16:10:13 <API> tuesdays?
16:10:17 <joanie> yup
16:10:32 <joanie> hopefully there will be some list discussion
16:10:46 <joanie> but my money (so to speak) is on it being an agenda item
16:10:51 <clown> so, Tue Nov 20 is the next meeting?
16:10:55 <joanie> yes
16:10:59 <clown> gotcha
16:12:07 <jjmarin> so then, presumably there won't be any update till that date :-/
16:13:09 <API> jjmarin, I think so
16:13:15 <API> so more questions, doubts?
16:13:16 <API> moving?
16:14:55 <joanie> sitting still?
16:14:58 <API> nobody asking so miving
16:15:00 <API> moving too
16:15:01 <clown> twiddling thumbs.
16:15:11 <API> GNOME 3.6.2 and/or 3.8 updates
16:15:13 <API> ups
16:15:16 <clown> (actually reading email — oh the guilt)
16:15:19 <API> #topic GNOME 3.6.2 and/or 3.8 updates
16:15:33 * joanie cheers for Python 3
16:15:48 <API> #info Finally dash support on orca was accepted, and included both on master and gnome-3-6 branch
16:16:04 <API> https://bugzilla.gnome.org/show_bug.cgi?id=686583
16:16:04 <tota11y> 04Bug 686583: normal, Normal, ---, gnome-shell-maint, RESOLVED FIXED, Regression: Dash item labels fail to report accessible names
16:16:19 <API> #info if nothing strange happens it will be available on 3.6.2 next week
16:16:35 <jjmarin> API: well done !
16:17:00 <clown> does "Dash" == "Dasher"?
16:17:12 <clown> or, what's Dash?
16:17:18 <API> #info finally I agreed with Rui Matos and Daiku to have a 3.6.2 release with the most critical stuff
16:17:26 <jjmarin> clown: dash is the gnome 3 dock
16:17:27 <API> clown, the dash is this column at the left
16:17:29 <API> at the overview
16:17:39 <clown> API, I"m on a Mac :-)
16:17:46 <API> some people call it the "Favourites"
16:17:54 <clown> But, I know what you mean.
16:17:58 <clown> coolness...
16:18:15 <API> joanie, for minutes sake
16:18:29 <API> s/Daiku/Daiki Ueno
16:18:40 <API> and done
16:18:40 <joanie> k
16:18:42 <API> next?
16:18:50 <clown> I have a 2 things...
16:19:30 <clown> #info Joseph explored the re-entrancy problem in gnome-shell and the JavaScript focus tracker.
16:19:50 <clown> #info he discovered a couple of other reasons that do not solve the problem.
16:20:11 <clown> #info He will summarize these findings in the relevant bugzilla
16:20:18 * clown looks up the bugzilla...
16:20:59 <clown> #info https://bugzilla.gnome.org/show_bug.cgi?id=681276
16:20:59 <tota11y> 04Bug 681276: normal, Normal, ---, liyuan, UNCONFIRMED, Reentrancy problems on gnome-shell.
16:21:22 <clown> that was the first one.  questions before I go on?
16:21:54 <joanie> I don't get your statement
16:22:01 <joanie> he discovered a couple of other reasons that do not solve the problem.
16:22:15 <clown> okay, let me clarify that.
16:23:05 <clown> background:  the problem occurs for any method call on the accessible in the event object.  E.g., accessible.get_name() causes the problem.
16:23:43 <clown> I thought that if I these methods were called in a deferred way, that might solve the problem.
16:24:22 <clown> I tried both an timeout_add() and and idle_add() on the main loop.
16:24:30 <clown> that did not solve the problem.
16:25:03 <clown> in a sense, I know where the problem does *not* lie.  Not as good as knowing where the problem does lie...
16:25:14 <joanie> aha
16:25:15 <clown> does that make things clearer?
16:25:16 <joanie> gotcha
16:25:19 <joanie> a bunch
16:25:20 <joanie> thanks!
16:25:32 <clown> you're welcome!
16:25:45 <API> well, in that case more that "discovered reasons" is more like "discarded hypothetical solutions"
16:25:47 <API> right?
16:25:58 <API> or discarded reasons
16:26:03 <clown> right
16:26:27 <clown> should I re-info that?
16:26:40 <joanie> sure
16:27:37 <clown> #info Joseph discounted hypothetical solutions to the problem by trying to execute the functions in a deferred (asynchronous) way using the main loop.
16:27:52 * clown really tried to use as much jargon as possible...
16:28:06 <joanie> :)
16:28:24 <clown> any other questions on this issue?
16:28:49 * jjmarin brainstorms this thought: AFAIK the notification API is going to be redesigned. I think it good idea to be around to be sure that this new API produce accessible notification
16:29:14 <jjmarin> sorry clown
16:29:25 <clown> no problem jjmarin.
16:29:34 <jjmarin> I only read other questions :-)
16:29:35 <clown> shall I go onto my second 3.6 related topic?
16:29:35 * API propose to talk about that notification thing at misc time
16:29:42 <API> clown, go on please
16:29:53 <clown> Okay, more background:
16:30:17 <clown> I was setting up a demo machine to show off the latest magnifier functions, and the interim focus tracking.
16:30:48 <clown> the focus tracking was only partially working — it worked fine for GNOME Shell widgets, and I believe with gedit.
16:31:03 <clown> But, when I tried FF or LibreOffice, nothing was tracked.
16:31:29 <clown> I went on to make sure that Orca was installed and working — I launched it and tried a few things.
16:31:43 <clown> Then I quit Orca, and went back to looking at the focus tracker.
16:31:58 <clown> Suddenly everything was working fine in that regard.
16:32:05 <joanie> what version of gnome?
16:32:25 <clown> It *looks* like Orca does some extra setup of accessibility?  Is that possible?
16:32:32 <joanie> not setup
16:32:38 <joanie> it checks if it is enabled
16:32:41 <joanie> and if not, it enables it
16:32:43 <clown> re:  Gnome version — I think it was 3.6.5
16:32:51 <joanie> there is no 3.6.5
16:32:57 <clown> I'd have to go get the demo machine and fire it up.
16:33:06 <clown> 3.6.3?
16:33:25 <clown> it was a F18 alpha with all the latest updates as of last Fri.
16:33:33 <joanie> 3.6.1
16:34:02 <joanie> perhaps accessibility always on isn't always on for Gecko or VCL
16:34:05 <joanie> is my guess
16:34:39 <API> yeah probably they didn't update that yet, and they are still using gnome2 procedures
16:35:05 <API> I didn't take a look to the last stuff they have
16:35:25 <API> anyway, if orca is running
16:35:29 <API> and after that you run ff
16:35:33 <API> that should work
16:35:36 <API> in that senes
16:35:44 <API> clown, when you made that focus tracking test
16:35:44 <clown> no, API, the sequence was:
16:35:50 <API> you were trying it with orca?
16:35:56 <joanie> clown: Orca doesn't turn OFF a11y
16:36:10 <API> in the sense of discard if were working with orca but not with the focus tracking?
16:36:20 <clown> 1. launch focus tracker.  2. launch FF, 3. launch Orca, 4. quit Orca.  Focus tracker works in FF.
16:36:32 <joanie> 16:35:55 <@joanie> clown: Orca doesn't turn OFF a11y
16:36:44 <joanie> It does turn it on if it's not already on
16:37:02 <API> clown, and what happened at 3.5?
16:37:13 <API> without quitting orca focus tracking was working or not?
16:37:58 <clown> API, to the best of my recollection, it always worked.  But, I was really looking that closely at it then.
16:38:31 <clown> btw, I went the machine.  it is 3.6.1
16:39:02 <clown> joanie, re:  it does turn it on if it's not already on.  Well, before launching orca, there was some focus tracking.
16:39:12 <joanie> in gtk
16:39:16 <mgorse> Ideally, Firefox and LibreOffice should handle a11y similarly to gtk (ie, have it always enabled and create accessibles lazily), so maybe it's worth filing bugs against them. For the time being, the caret focus tracker could enable the a11y setting if it isn't already enabled, to work around it
16:39:19 <joanie> because a11y is always on
16:39:31 <clown> Ah, I get it now.
16:39:38 <clown> thanks mgorse and joanie
16:39:54 <joanie> gnome-shell should also be fine
16:40:06 <clown> yes, joanie gs was fine.
16:40:14 <API> joanie, gnome-shell already made the move to use the last changes on a11y initialization
16:40:23 <joanie> i know
16:40:26 <joanie> this is why it should be fine
16:40:28 <joanie> ;)
16:40:30 <API> as I said not sure if ff did that
16:40:31 <joanie> it's the outsiders
16:40:41 <API> anyway, clown could you send a mail to the list
16:40:49 <API> probably with the last version of your focus tracking
16:40:55 <API> so anyone could test a little
16:41:10 <clown> so, another test would be, say, epiphany?  (is that the name of the insider web browser?)
16:41:12 <API> hopefully someone like alexander surkov
16:41:17 <API> or trevor would be listening
16:41:32 <API> clown, yes it would a good thing to test epiphany too
16:41:57 <clown> API,  that's a good start (an email to the list).  Which list?  a11y-devel?
16:42:05 <API> clown, yep
16:42:14 <clown> okay.
16:42:31 <API> clown, anything else?
16:42:34 <clown> that's all my "updates" with respect to 3.6/3.8.
16:43:01 <API> ok, thanks
16:43:07 * joanie has hers pretyped
16:43:12 <API> so anyone else wants to add any update ?
16:43:20 <joanie> #info An ugly Gtk+ crasher was found by a user and is now fixed by Cosimo (yay!) Should be included in GNOME 3.6.2. (See bug 687872.)
16:43:21 <tota11y> 04Bug https://bugzilla.gnome.org/show_bug.cgi?id=687872 major, Normal, ---, gtk-bugs, RESOLVED FIXED, Segfault when attempting to get character extents for GtkTextCellAccessible objects
16:43:23 <joanie> #info Joanie did a small-but-big performance improvement for Orca's flat review which will be included in GNOME 3.6.2.
16:43:26 <joanie> #info Python 3 has been approved GNOME 3.8.
16:43:29 <joanie> #info Orca's been ready since before the 3.6 release, though I'm still finding and fixing the occasional bug.
16:43:31 <joanie> #info Speech-dispatcher and liblouis seem to be ready to go.
16:43:34 <joanie> #info BrlApi is still being tested by Joanie. She submitted a tiny patch to the BrlTTY folks and is testing and bug fixing in Orca.
16:44:09 <joanie> jhernandez: when will you have your Python 3 work done?
16:44:49 <joanie> (it would also be worth testing in 3.3 as I am finding issues which were not present in 3.2)
16:44:55 <jhernandez> joanie: ASAP
16:45:22 <jhernandez> I was waiting for the acceptation as an official Goal
16:45:56 <joanie> jhernandez: many of the patches are python 2 safe
16:45:57 <jhernandez> it's on top of my accerciser's TODO list
16:46:04 <joanie> and clearly marked as such if memory serves me
16:46:58 <jhernandez> great!
16:48:13 <API> ok, someone else?
16:48:58 <API> ok, lets move
16:49:10 <API> #topic Closer ties with standards-writing groups (e.g. W3C)
16:49:12 <API> joanie?
16:49:27 <joanie> ok, she says looking at clown ;)
16:49:34 <joanie> so I've been thinking....
16:49:39 <clown> and what does she see?
16:49:40 * joanie infos
16:49:45 <joanie> (a red nose)
16:49:52 <clown> lol
16:50:00 * clown doesn't drink that much.
16:50:30 <joanie> #info Joanie has been thinking for a while now that it would be good if the standard-definers/writers and the standard-implementors interacted more so that they each knew what the other was doing.
16:51:02 <joanie> #info Joanie (and other developers here) lacks the time to be "members in good standing" in W3C groups.
16:51:19 <joanie> #info Joanie would like to find some other way to coordinate better and share information.
16:51:29 <joanie> #info Joanie notes that Joseph is here each week. ;)
16:51:48 <joanie> #info and he serves on several(?) W3C groups.
16:51:57 <clown> two
16:52:06 <jjmarin> wow
16:52:18 <joanie> So I'm wondering if perhaps like the weekly marketing updates, we have weekly W3C updates
16:52:30 <joanie> that would "force" (air quotes) us to talk each week about this stuff
16:52:37 <joanie> so that we keep each other in the loop
16:53:39 <clown> I think the idea of the standards groups working with implementors is a great idea.
16:53:51 <joanie> (yay)
16:54:51 <clown> I know that recently working with mozilla (Alex and David) regarding ARIA has resulted in some surprises.
16:55:07 <clown> And the main complaint is that the W3C is not transparent enough.
16:55:30 <API> clown, so some info to include today, or you are still recovering from the surprise?
16:55:36 <clown> note that David is a member of the ARIA group as ell.
16:55:39 <clown> "well".
16:56:05 <jjmarin> Does he surprise himself ? :-)
16:56:07 <clown> API, I'm just not sure what to report on.
16:56:18 <clown> jjmarin, every time he looks in a mirror.
16:56:27 <jjmarin> :-)
16:56:39 <joanie> report on each week you mean?
16:56:39 <clown> let's try this...
16:56:55 <jjmarin> clown: In which groups are you ?
16:56:55 <API> well, as joanie said, some kind of short note about what happened on last meeting
16:57:09 <clown> no joanie.  API was suggesting "some info to include today".
16:57:18 <joanie> aha
16:57:20 <clown> okay, here. goes.
16:57:59 <clown> #info the last ARIA meeting involved a group (of about 4 participants) working through the set of test cases.
16:58:23 <clown> #info the point was to determine if the test cases were valid, fix them if there were not, and execute the test.
16:59:15 <clown> #info "execute the test" means loading the test file into a browser, and running an accessibility inspector to see if the predicted result matched the actual a11y tree.
16:59:42 <clown> #info browers include FF, IE, Chrome, and Safari.
17:00:07 <clown> #info (detail) Josepoh is concentrating testing FF on GNOME. (AT-SPI).
17:00:20 <clown> s/Josepoh/Joseph.
17:00:29 <clown> s/Josepoh/Joseph/
17:00:53 * API wonders if would make sense to propose to include epiphany on the tests ...
17:01:15 <joanie> latest unstable as well since API and I are hacking on it
17:01:36 <joanie> we could see what's missing in the implementation
17:01:39 <API> I also wonder what kind of tests
17:01:47 <API> that test file is an html file I guess
17:02:08 <clown> makes sense, but we need a tester (and I would have to lobby for that person to have access to the test cases).
17:02:30 <clown> API, here is the structure:
17:02:46 <joanie> which gets back to my belief that there is a wall between those determining standards and those implementing them
17:02:47 <clown> A test case includes a testable statement, and expected result, and a test file.
17:03:14 <clown> the test file is html, and is representative of the testable statement.
17:04:58 <API> well it is similar to what webkit have
17:05:04 <clown> You get around the wall partially by joining the W3C.  So, businesses (IBM, Apple, Google), and Universities and other stake holders send representatives.
17:05:06 <API> webkit has thousands of tests
17:06:43 <clown> I wonder how many "GNOME" people have joined the W3C.
17:07:05 <clown> they do constitute a stake holder.
17:07:17 <API> clown, well most businesses related to GNOME are small companies, except Red Hat
17:07:35 <clown> API, you don't have to be a business.
17:08:44 <clown> actually, Janina is a member.
17:09:00 <joanie> I think I'll raise this with the Board
17:09:16 <joanie> #action Joanie will talk to the Board about GNOME joining the W3C
17:09:22 <clown> Janina is somehow associated with GNOME.
17:09:30 <joanie> really??
17:09:42 <clown> yes, but I'm not sure how.
17:09:54 * joanie grins
17:09:58 <clown> her email is 'rednote.net'.  Is that clue?
17:10:19 <joanie> I love fedora but fedora ain't gnome if that is your point
17:10:45 <joanie> anyway, my action item still stands
17:10:46 <clown> turns out rednote.net is her own web site.
17:11:06 <joanie> anyway, deep dive
17:11:16 <clown> no, I had no point about fedora.
17:11:18 <joanie> clown: so you are cool with a weekly agenda item?
17:11:30 <clown> sure.
17:11:30 <joanie> so test access or not, we can stay in the loop?
17:11:35 <joanie> thanks!
17:12:02 <clown> is what I info'ed above the right kind of detail?
17:12:09 <joanie> yup
17:12:15 <joanie> depending on the topic
17:12:39 <joanie> if there are specific ARIA things worth looking at (e.g. in WebKitGtk) I'd love to know what they are
17:12:43 <joanie> if that is not super-secret
17:12:57 <joanie> pending my getting GNOME as a member (assuming I can)
17:13:44 <clown> okay.
17:13:46 <joanie> otherwise, there will just be more suprises and/or standards that are not implemented
17:14:10 <clown> for the record:
17:14:29 <joanie> and no feedback about alternatives (i.e. personally, I think there are many things about the Gecko a11y implementation that are not ideal)
17:14:30 <clown> a W3C standard is not declared a standard until it can show it has at least two implementations.
17:14:39 <joanie> right
17:14:49 <joanie> and then it becomes a standard
17:14:55 <joanie> and then officially it trickles down to us
17:15:02 <joanie> and then we say "wtf, why??"
17:15:08 <clown> and, that's what the current testing effort is about — showing that ARIA has at least two implementations in browsers.
17:15:11 <joanie> and by then it's too late
17:15:19 <jjmarin> I guess marketing team  would love to know that GNOME join W3C :-)
17:15:28 <joanie> we haven't joined yet
17:15:41 <jjmarin> just in case I mean
17:15:54 <joanie> if we join, Karen will no doubt tell the marketing team
17:16:01 <jjmarin> sure
17:16:06 <clown> another tidbit of into, for those who don't know.  W3C works by consensus.
17:16:31 <joanie> of the invited
17:16:55 <API> so ... taking into account that we all agreed here, and we are over time ... moving to next topic?
17:17:03 * joanie nods
17:17:16 <API> so moving
17:17:27 <API> #topic Marketing
17:17:29 <API> jjmarin, ?
17:19:41 <jjmarin> #info Juanjo has started a thread in the gnome accessibility mailing list to discuss about how to update the accessibility information from the GNOME web pages. https://mail.gnome.org/archives/gnome-accessibility-list/2012-November/msg00003.html
17:20:04 <jjmarin> #info Juanjo will write an article for the GNOME annual 2011 about the FoG Accessibility Campaign
17:20:07 <jjmarin> done !
17:20:16 <jjmarin> questions ?
17:21:25 <clown> jjmarin, did you see Jan's reply regading the documentation I wrote on the zoom options control center?
17:22:01 <jjmarin> I´ve got the time to read the email yet
17:22:44 <clown> okay.  my question was: is this along the lines of what you are looking for?
17:22:55 <clown> let me know when you've had a chance to read it.
17:23:38 <jjmarin> no, it wasn´t. I guess we can include a section about the zoom options in the User Guide
17:24:16 <clown> okay
17:24:41 <API> so moving?
17:25:18 <API> #topic Reminder: Q3's
17:25:20 <API> this is easy
17:25:35 <API> #info please update your Q3 report with your information
17:25:36 <API> thanks
17:25:44 <API> #topic misc time
17:25:52 * API rushing as we are 25 over time
17:25:58 <API> something to add not scheduled?
17:26:00 <clown> regarding Janina's affiliation:  it's not GNOME.  It's the Linux Foundation.  (Slightly broader, I suppose).
17:26:05 <clown> she lists herself as "Chair, Accessibility Workgroup, Free Standards Group", but the link goes to:
17:26:11 <clown> http://www.linuxfoundation.org/collaborate/workgroups/accessibility
17:26:13 <clown> (done).
17:28:54 <clown> didn't jjmarin have a brainstorm earlier?
17:28:57 <clown> ah, he left...
17:29:05 <clown> never mind :-)
17:29:20 * joanie pokes API
17:29:30 <API> as nobody else added anything
17:29:37 <API> it is a good moment to close the meeting
17:29:40 <API> thanks everybody
17:29:44 <clown> thanks API
17:30:15 <API> sorry if someone of you have a bad surprise on these 30 minutes extra
17:30:17 <API> was not planned :P
17:30:19 <API> #endmeeting