16:07:51 <API> #startmeeting
16:07:51 <tota11y> Meeting started Thu Jul 11 16:07:51 2013 CET.  The chair is API. Information about MeetBot at http://wiki.debian.org/MeetBot.
16:07:51 <tota11y> Useful Commands: #action #agreed #help #info #idea #link #topic.
16:07:59 <API> #topic GSoC updates
16:08:18 <API> hmm, is not changing the topic
16:08:26 <API> joanie, I guess that tota11y is dead
16:08:30 <joanie> 16:06:10 <joanie> bot lost ops
16:08:36 <joanie> 16:06:19  >>> joanie gets ops back
16:08:56 <clown> should I wait until ops is back?
16:09:01 <magpie_> what's ops?
16:09:13 <clown> great power to API
16:09:21 <API> magpie_, grants to do stuff
16:09:29 <clown> becomes king of the chat room.
16:09:33 <API> without ops tota11y can't change the chat room
16:09:36 <API> in any case
16:09:38 <magpie_> channel operator?
16:09:45 <API> clown, magpie_ I think that you can go on
16:09:57 <clown> ok, I will proceed.
16:10:10 <clown> #info Magdalen put together a patch for the FocusCaretTracker, and uploaded it to the bugzilla. (Congratulations).
16:10:24 <clown> #info Jasper reviewed it, and made some interesting points.
16:10:29 <joanie> :)
16:10:43 <clown> #info Joseph responded to his review, and has been ruminating over it since.
16:11:02 <clown> #info Joseph believes the way to address them is to answer a few technical questions:
16:11:13 <clown> #info 1. should the tracker be a singleton?
16:11:18 <magpie_> #link https://bugzilla.gnome.org/show_bug.cgi?id=647074
16:11:18 <tota11y> 04Bug 647074: enhancement, Normal, ---, gnome-shell-maint, UNCONFIRMED, GNOME Shell Magnifier should track focus and the caret
16:11:34 <clown> #info 2.  where should Atspi.init() be done?  (probably in main.c).
16:11:52 <clown> just a second, I have to look up my notes.
16:12:22 <clown> #info 3.  should event registration with Atpsi happen in the constructor?
16:12:58 <API> clown, in the constructor of what?
16:13:04 <clown> #info 4. (possibly a 4th question)  should clients have the opporunity to respond to focus event separate from slection events?
16:13:04 <API> of the tracker?
16:13:09 <clown> API, yes.
16:13:15 <API> any other  ojbect of gnome-shell?
16:13:15 <API> ok
16:13:24 <magpie_> #info Magdalen sent a copy of APIs advice chat log to the mentors
16:13:30 <magpie_> this morning
16:13:56 <API> magpie_, could you elaborate that?
16:14:02 <API> a brief summary of that chat log?
16:14:07 <clown> #info while these issues are discussed with the "experts", Joseph feels that Magdalen should proceed with changing the magnifier to use the tracker as is.
16:14:08 <magpie_> #info at the moment it's in the constructor since update
16:14:13 <magpie_> #link https://github.com/thisMagpie/GSoC/blob/master/javascript/focusCaretTracker.js
16:14:31 <magpie_> API you remember when you came to advice us?
16:14:31 <clown> #info Joseph feels that Magdalen should leave the tracker alone for the time being.
16:14:39 <magpie_> 6th june
16:14:52 <API> yes, I remember a conversation between you, Jasper and me
16:15:00 * clown wonders why he's bothering with the infos, when another conversation just starts up one its own,..
16:15:07 <API> clown, sorry
16:15:12 <API> as magpie_ started with her infos
16:15:17 <API> I thought that you had finished
16:15:26 <API> in any case you are right
16:15:35 <clown> I'll say done — quetions, when I'm done, okay?
16:15:37 <magpie_> API don't blame me!
16:15:38 <API> please a little order first
16:15:52 <API> magpie_, sorry, but clown was right
16:15:55 <API> he was talking
16:16:00 <API> and he didn't finished yet
16:16:06 <magpie_> sorry clown and API
16:16:12 <clown> s'okay.
16:16:21 * clown collects his thoughts.
16:16:23 <API> clown, as I said sorry as well
16:16:48 * API stopping talk until clown says the 'done' word
16:17:16 <clown> #info Magdalen sent (as she info'ed) an email showing an IRC log of the converstaion with API about where Atspi should be initialized (question 1. above).
16:17:24 <clown> #info Joseph responded to that this morning.
16:17:51 <magpie_> oh
16:17:54 <clown> #info Joseph is throwing the question to the Atpsi experts:  where is the best place to do this?
16:18:24 <clown> #info Regarding GSoC "adminstration":  Joseph is going on a two week vacation at the beginning of August:
16:18:47 <clown> #info if there is any GSoC stuff he must do, that should be discussed soon.
16:19:07 <clown> #info Also, there should be a plan for some of the other mentors to take over in his absence.
16:19:18 <clown> whew!  done.  Questions?
16:19:44 <magpie_> I've aa couple more
16:19:52 <magpie_> infos
16:19:56 <API> clown, well, in fact you were the one that made a question
16:20:00 <API> oh. sorry
16:20:10 <API> magpie_, please go on, we can debate when you both finish
16:20:22 <magpie_> questions/or infos API?
16:20:28 <magpie_> ok
16:21:23 <magpie_> #info Magdalen wrote another blog post but it was not very detailed so thinks she should talk more about the gjs work in shell soon
16:21:57 <magpie_> #info Magdalen has got further on the magnifier
16:22:35 <magpie_> #info Magdalen also uploaded the work to github with the focuscarettracker
16:24:07 <magpie_> #info Magdalen is not sure how to change from global.get_pointer() without changing a lot of the mouse codes in magnifier
16:24:10 <magpie_> ok done
16:24:16 <clown> s/question 1. above/question 2. above/
16:24:22 <API> clown, well as I said
16:24:31 <API> in fact you were the one doing a question here
16:24:35 <API> <clown> #info Joseph is throwing the question to the Atpsi experts:  where is the best place to do this?
16:24:43 <clown> four questions, API (at least).
16:24:57 <magpie_> sorry that took long to get out
16:25:00 <API> yes, but probably the others can't be replied here
16:25:15 <API> at least
16:25:18 <clown> oh, you have reply to that one, API?
16:25:20 <clown> cool.
16:25:24 <API> questions like: "if should be a singleton"
16:25:32 <API> seems more like a question to gnome-shell developers
16:25:40 <API> but probably I'm wrong
16:25:53 <API> that specific question was specifically oriented to people in this room
16:26:03 <clown> there is an Atspi aspect to the singleton question.
16:26:04 <joanie> the atspi question *might* be something for mgorse to chime in on?
16:26:18 <API> clown, or I am wrong and all those questions were targeted for everybody here?
16:26:40 <clown> the Atspi aspect is:  with a singleton there is only one registration for the focus/caret events.
16:26:44 <magpie_> Jasper said he prefer not to be in a constructor already you see
16:27:23 <clown> API, these questions were targeted to everyone here.  And, mgorse's opinion is a really good place to go.
16:28:15 <mgorse> Which events is the code listening for? Just focus events?
16:28:31 <magpie_> focus caret and selected mgorse
16:28:48 <clown> mgorse:  focus, select, and caret  (right Magpie)
16:29:16 * clown looks to see exactly which events.
16:29:48 <magpie_> https://github.com/thisMagpie/GSoC/blob/master/javascript/focusCaretTracker.js
16:30:19 <clown> mgorse:  object:state-changed:focused, object:state-changed:selected, and object:text-caret-moved
16:30:51 <mgorse> I think I'll need to test that AT-SPI works correctly with multiple listeners listening for the same event. Sorry--I should have done this before now.
16:31:19 <clown> mgorse, no problem.  And this is exactly the kind of information I was looking for.  Thanks!
16:31:46 <clown> mgorse, could you make a note of that on the bugzilla?
16:31:51 <joanie> mgorse: when I did the proof-of-concept magnification python listener, it and orca were each listening for those events
16:31:54 * clown looks up bugzilla.
16:31:56 <joanie> and it worked
16:32:00 <mgorse> I don't think it has any specific code to, say, not send a D-Bus match rule if one has already been sent for the event. I'm not sure off-hand if dbus-daemon will send the event twice if it has two match rules for the event
16:32:03 <joanie> mind you, that was what.... a couple of years ago?
16:32:16 <clown> https://bugzilla.gnome.org/show_bug.cgi?id=647074#c18
16:32:17 <tota11y> 04Bug 647074: enhancement, Normal, ---, gnome-shell-maint, UNCONFIRMED, GNOME Shell Magnifier should track focus and the caret
16:32:57 <clown> joanie:   the POC was a singleton, wasn't it?
16:33:04 <joanie> POC?
16:33:12 <magpie_> proof of concept
16:33:12 <clown> proof-of-concept
16:33:15 <mgorse> joanie: Your listener was separate from Orca, right? Ie, in a different process? I should check that it doesn't, say, send the event multiple times to each listener if there are multiple listeners in the same process
16:33:20 <joanie> mgorse: yes
16:33:27 <joanie> it was totally separate/independent
16:33:29 <magpie_> it never worked for me
16:33:52 <clown> mgorse, good point, because the tracker clients will all be in the same gnome-shell process.
16:34:55 <clown> mgorse, can you make a note of that on the bugzilla — the tests you are going to carry out?  and/or the results?
16:35:33 <mgorse> yeah, I need to comment on the bug anyhow
16:35:44 <clown> mgorse, much appreciated.
16:35:50 <mgorse> #action mgorse will test AT-SPI's behavior when registering multiple event listeners in the same process and comment on the bug
16:35:55 <magpie_> thanks mgorse
16:36:24 <clown> so:  the tracker code changes should wait until some of these issues are resolved.
16:36:26 <clown> IMHO
16:36:45 <clown> s/some of these/all of these/
16:36:55 <magpie_> well i'm far as i can get with both files
16:37:18 <magpie_> i have to find out about global.get_pointer from clutter
16:37:37 <clown> why?  well, we should take that out of the meeting.
16:38:17 <magpie_> sure i was updating on my progress with the code that's all
16:39:07 <clown> API, do you have questions about any other of my other questions?
16:39:10 <API> clown, btw, as mgorse
16:39:17 <API> also mentioned the "comment thing"
16:39:24 <API> I will comment there about the need of init atspi
16:39:30 <clown> thanks API.
16:39:34 <API> sorry, I forgot to do that before
16:39:47 <magpie_> thanks API
16:39:55 <API> in any case
16:40:03 <API> as some of the questions are still in the air
16:40:14 <API> what about send them to the accessibility dev mailing list?
16:40:35 <clown> Sure — send them all?
16:41:35 <clown> #action Joseph will summarize the 4 main questions about the FocusCaretTracker, and send that to the accessibility dev list.
16:41:52 <API> just saying because as you said before
16:41:58 <API> there are stuff to investigate
16:42:04 <clown> makes sense.
16:42:16 <API> and probably we will not have time to solve that on ~#11y-meeting, or on #a11y today
16:42:28 <API> so with that action, I think that we can move on
16:42:29 <clown> I know I don't have the time :-(
16:42:53 <API> clown, fwiw, thanks for coming to the a11y-meeting and do the summaries of situation
16:43:04 <magpie_> #action Magdalen will email the list
16:43:11 <clown> my pleasure:  magnifier users need focus tracking....
16:43:19 <clown> I feel for them.
16:44:32 <API> hmm
16:45:02 <API> magpie_ a mail about what?
16:45:23 <magpie_> API clown said he doesn't have time so i don't mind doing it
16:45:37 <clown> magpie_ I have time to do the email.
16:46:01 <clown> I was saying that I don't have a lot of time to solve these problems in #a11y today.
16:46:08 <clown> I was agreeing with API.
16:46:11 <magpie_> oh i just read that <clown> I know I don't have the time :-(
16:46:25 <API> ok, so after this clarification
16:46:31 <API> magpie_ thanks for your offer
16:46:41 <magpie_> sorry i thought u meant that not generally
16:46:42 <API> clown thanks for the (future) mail
16:46:48 <API> having said so
16:46:55 <clown> s'okay magpie_
16:46:57 <API> lets move, although I think that the rest of topic are lighter
16:47:37 <API> #topic W3C updates
16:47:48 <API> this is a just in case topic
16:47:49 <clown> well, hello there.  me again...
16:47:55 <API> yes
16:47:58 <API> clown, please go on
16:48:30 <clown> #info The ARIA working group desires to put together a "Last Call Draft" of the User Agent Implementation Guide, by the end of the month (July).
16:48:56 <clown> #info As editor, I will be making the final edits in preparation for that publication.
16:49:09 <clown> #info Hopefull, there will be a Last Call Draft by August.
16:49:20 <clown> s/Hopefull/Hopefully/
16:49:24 <joanie> that would be nice
16:49:37 <joanie> I'm hoping to have some breathing room after guadec
16:49:46 <clown> #info Then, the public is invited to make comments on that draft, and request changes, clarifications, etc.
16:49:59 <joanie> and would love to take that draft and look at it specifically in the context of webkitgtk
16:50:03 <clown> done.  questions?
16:50:21 <clown> that's an atspi platform, right?
16:50:28 <joanie> ATK
16:50:32 <joanie> but yeah
16:50:35 <joanie> gnomey
16:50:35 <API> joanie, but clown is right
16:50:38 <clown> yeah, it's confusing...
16:50:41 <API> for the webkit2 gtk part
16:50:44 <API> at-spi2 is needed
16:50:50 <joanie> really?
16:50:50 <clown> the UAIG says ATK/AT-SPI.
16:51:00 <API> joanie, due the plug/socket related tests
16:51:07 <API> in fact on the internal jhbuild of webkitgtk
16:51:11 <API> they had atspi before atk
16:51:12 <clown> ATK since the user agent implements things using ATK, but the ATs look at what's coming out of AT-SPI.
16:51:17 <joanie> but the user agent implementation
16:51:19 <joanie> is done in ATK
16:51:29 <API> yes, but if you want to test it
16:51:35 <API> as they are in different processes
16:51:43 <API> and that works thanks to the atksocket/plug
16:51:45 <joanie> so we will not in WebCore Accessibility support for Gtk have any AT-SPI code
16:51:48 <joanie> right?
16:51:49 <API> parts that are implemented at atspi
16:51:56 <API> joanie, right
16:51:58 <joanie> k
16:51:59 <joanie> phew
16:52:03 <API> there are no atspi code there
16:52:15 <API> but in order to work, so in order to test some of the tests
16:52:18 <API> atspi is needed
16:52:19 <joanie> of course
16:53:01 <joanie> my interest is to be sure that the draft guidelines are fully doable and consistent with what we'd expect
16:53:05 <joanie> and we've just not had that time
16:53:18 <joanie> but if the draft is out in august and we are less-crazy-busy in august
16:53:21 <joanie> that would be perfect
16:53:40 <clown> the draft guidelines are being tested now (when I get a chance).  But, tested with FF.
16:54:00 <joanie> they'll just cram it all into AtkObject attributes :P
16:54:07 <joanie> I'd like to do it right ;)
16:54:12 <API> clown, just out of curiosity
16:54:16 <API> tested with FF on windows?
16:54:35 <clown> API, I'm testing on FF/Fedora 19.
16:54:57 <clown> API, Rich is testing FF/Windows.  Others are testing IE, Safari, and Chrome.
16:55:06 <API> clown, on thanks
16:55:20 <clown> you're welcome API.
16:55:55 <clown> just a reminder:  https://www.w3.org/WAI/PF/testharness/testreport?testsuite_id=1
16:56:06 <clown> that shows the platforms being tested along the top.
16:57:32 * clown not sure how accessilbe that table is, sorry.
16:57:52 <API> clown, ok, again thanks
16:58:03 <clown> you're welcome.
16:58:05 <API> so, more questions, doubts, comments?
16:59:52 <API> it seems that no
16:59:56 <magpie_> none from me
17:00:01 <API> I will skip marketing as jjmarin is not here
17:00:07 <API> so just in case
17:00:11 <API> #miscellaneous time
17:00:16 <API> argh
17:00:20 <API> #topic miscellaneous time
17:00:32 <API> something not scheduled (and really short) to add ?
17:01:18 <magpie_> i've offered to volunteer for guadec
17:01:25 <magpie_> and put my name down
17:02:00 <API> magpie_, well, have fun ;)
17:02:12 <clown> API, that was faintly ominous.
17:02:43 <API> clown, you are a evil thinker
17:02:52 <API> I was totally sincere ;)
17:02:54 <API> having said so
17:03:06 <API> I think that this is a good point to close the meeting
17:03:11 <API> thanks everybody for coming here
17:03:27 <API> see you on #a11y, and #a11y-meeting next week
17:03:30 <API> #endmeeting