16:00:00 <API> #startmeeting
16:00:00 <tota11y> Meeting started Thu Jan 16 16:00:00 2014 CET.  The chair is API. Information about MeetBot at http://wiki.debian.org/MeetBot.
16:00:00 <tota11y> Useful Commands: #action #agreed #help #info #idea #link #topic.
16:00:15 <API> #topic on 5 minutes of margin; waiting people
16:00:54 * clown hums
16:03:04 * joanie grumbles at Mozilla
16:03:17 * clown notes that the agenda is out of date.
16:03:26 <joanie> clown: my bad
16:03:30 <clown> https://wiki.gnome.org/Accessibility/Meetings#Agenda_for_the_Next_Meeting_.289_January.29
16:03:31 <joanie> but it's the same ol' agenda
16:03:42 <joanie> just the date
16:03:46 <joanie> oh noes!
16:03:47 <clown> yes, that's was me second thought.
16:03:48 <joanie> :P
16:04:04 <clown> why are you grumbing at Mozilla?
16:04:22 <clown> dare I ask?
16:04:50 <joanie> see #a11y
16:05:06 <joanie> "not sure how Windows screen readers do here but they manage it somehow. I can't change the behavior on Windows and I would try to avoid to make the change for ATK since it's sort of big difference."
16:05:10 <joanie> that's a Surkov quote
16:05:38 <joanie> since Windows wants it their way, emitting a single, lousy state-changed:focused event for a focsable widget is apparently a "big deal"
16:05:44 <joanie> ENDOFRANT
16:06:14 <clown> are they at least emitting a focus event?  (I don't have a log from far enough back in a11y)
16:06:54 <joanie> focus event is deprecated
16:06:57 <joanie> but I don't think they are
16:07:04 <joanie> and they are not updating the state of the listbox
16:07:08 <joanie> anyhoo, it's meeting time
16:07:10 * clown assumes we have test cases in the aria test suite that tested something like this, but state-change:focus wasn't required back then.
16:07:22 <joanie> this isn't even aria
16:07:34 <joanie> it's a fricken <select>
16:07:44 <joanie> anyhoo, I'll shut up now
16:07:57 <clown> but, if there was an aria listbox with selectable children that did what you wanted, wouldn't you want to know?
16:08:07 * clown shuts up too.
16:08:10 <joanie> :)
16:08:51 <API> well, just waiting for the dialog end
16:08:54 <API> I guess that ended
16:09:02 <API> so starting (for real) the meeting
16:09:10 <joanie> The Joanie and clown show might go on forever :P
16:09:13 <joanie> thanks API
16:09:18 <API> #topic Focus-tracking deprecation
16:09:38 <API> #info API has still pending that mail to gtk maintainers about retaking this
16:09:47 <API> sorry, you can throw the stones at the end of the meeting
16:09:52 <joanie> nah
16:09:57 <API> other comments, doubts, opinions?
16:10:06 <joanie> just a slight one about the previous discussion
16:10:36 <clown> what if the stones are made from styrofoam?
16:11:05 <joanie> #info Based on mozilla bug 960241 and associated discussion, Joanie is starting to think that Matthias is correct. What *exactly* the "spec" should look like remains up in the air. But being able to point to concrete ATK docs that clearly state expectations would be handy.
16:11:05 <tota11y> 04Bug https://bugzilla.mozilla.org/show_bug.cgi?id=960241 normal, --, ---, nobody, NEW , Tabbing into a list with selectable children should cause the list to emit an accessible state-changed:focused event
16:11:18 <joanie> (done)
16:11:54 <API> could you summarize that "matthias is correct" thing?
16:11:54 * joanie mutters "Why is "has keyboard focus" such a confusing thing?
16:12:14 <API> I mean
16:12:15 * clown mutters it has something to do with active descendant, maybe.
16:12:16 <API> is correct on what?
16:12:33 <joanie> API what I mean is, Matthias is failing to implement what I think should be pretty straightforward
16:13:05 <joanie> But he said (abbreviated version colored by my attitude): No. Not until we have a "spec" that defines this stuff.
16:13:36 <joanie> The fact that currently we get a "focus:" event for a focused object suggests to me that emitting object:state-changed:focused is a no-brainer
16:13:42 <joanie> but apparently I was wrong
16:14:14 <API> well, after my experience with key-focus-stuff on clutter
16:14:15 <clown> a question:  the issues revolves around accessibility events, correct?  Which may or may not be the same as GUI events.
16:14:17 <joanie> API so he's correct in that we need to figure out what the darned spec needs to say to convince implementors to do it
16:14:28 <API> key-focus-navigation can be classified as straightforward
16:14:31 <API> anyway
16:14:34 <API> thanks for the summary
16:14:46 <API> s/can be/can't be
16:14:48 <joanie> clown: yes, and I know. It's another reason we "need the spec"
16:15:04 <joanie> my point is I think we may need to prioritize the spec
16:15:10 <API> ok, makes sense
16:15:13 <clown> well, I wonder if the GUI event system is documented as good as it should be.
16:15:26 <joanie> whose gui?
16:15:40 <clown> GTK, I presume, in this case.
16:15:50 <API> nope
16:15:50 <API> :P
16:15:51 <joanie> oh, I don't think it is, no
16:15:55 <joanie> but that's not the point
16:16:09 <joanie> and if we get into a pissing match a la "your spec sucks too!"
16:16:17 <clown> so the GTK event documentation sucks as bad as the a11y event documenation?
16:16:20 <joanie> I still won't get what I want, namely the expected a11y events
16:16:26 <joanie> imho it does
16:16:28 <API> clown, more or less
16:16:34 <API> but the difference is that no one is expected
16:16:38 <API> to implement gtk
16:16:42 <clown> no, I was going to suggest that if there is *a* spec that is "good", use it as a starting point.
16:17:01 <joanie> clown: yeah, I wanted to point to it in response to matthias
16:17:06 <joanie> and found it sucked as well
16:17:12 <joanie> anyhoo, I'll rant until we move on
16:17:13 <clown> *sighs*
16:17:16 * joanie tries to reshut up
16:17:17 <joanie> :)
16:17:24 <API> so moving to next topic?
16:17:31 <jjmarin> hi
16:17:40 * clown wonders if the java event system documentation is any good.  Including javax.accessibility package.
16:17:46 <jjmarin> I think it would nice to fix https://bugzilla.gnome.org/show_bug.cgi?id=608231
16:17:46 <tota11y> 04Bug 608231: enhancement, Normal, ---, control-center-maint, UNCONFIRMED, Large Cursor For Low Vision
16:17:47 <clown> sure, move
16:17:53 <joanie> dunno about the docs clown
16:18:02 <joanie> but the implementation is 1000 kinds of broken
16:18:03 <joanie> ;)
16:18:22 <joanie> shutting up.... NOW!
16:18:23 <API> jjmarin, hmm, yes probably, but my square-mind is screaming "not related with the current topic" :P
16:18:23 <jjmarin> sorry clown
16:18:28 <clown> hi jjmarin!
16:18:47 <jjmarin> sorry, I think it was about gnome 3.12 features
16:18:49 <API> jjmarin, we can retake that at topic 3
16:19:01 <API> jjmarin, current topic is "Focus-tracking deprecation"
16:19:17 <API> anyway moving
16:19:17 <jjmarin> :-)
16:19:24 <API> #topic Wayland
16:19:44 <API> #info API also need to send a email resuming the discussion at the wayland-dev list
16:19:49 <API> more reasons for those stones
16:19:58 <API> anyway, probably it is worthy to retake that
16:20:04 <API> if we have something new to say
16:20:07 * clown tosses a few pebbles half-heartedly.
16:20:10 <API> mgorse, any luck about the
16:20:28 <API> security stuff for at-spi2?
16:21:06 <mgorse> It looks like dbus's authentication mechanism involves being able to read a cookie in the user's home directory
16:21:35 <mgorse> which I guess wouldn't be helpful in terms of doing what we need, since, as I understand it, being able to authenticate based on the user isn't enough
16:22:13 <mgorse> regardless, the registry could have some mechanism to only allow a caller to move the mouse if it has passed some test
16:22:21 <mgorse> but I think the question is what that test should be, and I'm not sure
16:22:46 <API> well, but we are also thinking on snooping key events
16:23:03 <API> in any case
16:23:04 <mgorse> since AT-SPI would need to have a list of clients that are allowed to call these functions, presumably, and it needs to ensure that clients are what they say they are
16:23:17 <mgorse> yeah
16:23:25 <API> hmm
16:23:27 <API> so then
16:23:30 <API> for example
16:23:36 <API> as far as I thought
16:23:44 <API> stuff like installing new packages
16:23:49 <API> on modern desktop
16:24:02 <API> ask you for the password in order to continue
16:24:12 <API> I thought that those services were also using DBUS
16:24:18 <API> mgorse, am I wrong?
16:25:12 <mgorse> I'm not sure. I think they're probably using policykit, but I'm just guessing, and maybe policykit uses D-Bus internally
16:26:15 <API> ok, in any case you only had one week to take a look to this
16:26:37 <API> so if you don't mind, keep looking to DBUS authentication
16:27:26 <API> so, anything else in this point? comments, thoughts?
16:29:02 <mgorse> I think the main question is how to ensure that a client is trusted, and what that means. Ie, it sound as though "something owned by the current user" isn't sufficiently secure
16:29:48 <API> well, for me
16:30:01 <API> is important to move the "who decides is a trusted client"
16:30:03 <API> to at-spi2
16:30:13 <API> because if is the compositor the one
16:30:16 <API> that will decide it
16:30:29 <API> I foresee that will be complex to add new ATs in the future
16:30:37 <API> well, when I say "add new ATs"
16:30:45 <API> I mean "compositor trusting a new AT"
16:31:17 <API> the idea of this, is ensure that at-spi2 just don't allow anyone to use their services
16:31:30 <API> so it could be considered as a trusted service by the compositor
16:31:39 <API> by any compositor, btw
16:31:59 <API> because I guess that this would also apply no any non-gnome-shell but wayland-compliant compositor
16:32:08 <API> I know that kwin is also doing some wayland stuff
16:32:15 <API> not sure in which stage they are
16:32:23 <API> so, having said so
16:32:30 <API> anything else?
16:32:32 <API> moving?
16:32:35 <mgorse> ok
16:33:48 <API> #topic Other progress towards 3.12
16:34:19 <API> #info API was working on some minor clutter-bugs related to remove text (ie: alt+f2 dialog) pointed by joanie
16:34:20 <API> done
16:34:35 <joanie> #info Joanie is still doing the Great Orca Rewrite
16:34:36 <joanie> done
16:34:46 <clown> the GOR?
16:34:57 <joanie> :)
16:34:59 <jjmarin> hehe
16:35:02 <joanie> it's GORy
16:35:10 <mgorse> #info mgorse was commenting on some evolution and related bugs, hoping it would be useful, but sometimes it doesn't do any good to just write a comment.
16:35:10 <mgorse> done
16:35:10 <jjmarin> Lol
16:35:22 <joanie> mgorse: details?
16:35:25 <clown> :-)
16:35:28 <mgorse> I half-way want to just commit the patch on bug 669441
16:35:28 <tota11y> 04Bug https://bugzilla.gnome.org/show_bug.cgi?id=669441 critical, Normal, ---, gtkhtml-maintainers, NEW, Accessibility no longer works with gtk+ 3.2 and above
16:35:43 <joanie> mgorse: ping rtcm in #a11y
16:35:54 <joanie> oh what, that's gtkhtml
16:35:59 <mgorse> Why? Is he the maintainer?
16:36:01 <joanie> s/what/wait/
16:36:10 <mgorse> module seems unmaintained, possibly
16:36:13 <joanie> mgorse: no, but evo is something he's now managing I think
16:36:22 <joanie> but at first I didn't see the gtkhtml bit
16:36:31 <mgorse> oh. Good to know. I should maybe talk to him anyhow
16:36:35 * joanie nods
16:36:41 <joanie> I think he cares and is concerned
16:36:50 <mgorse> and I was hoping that someone would know how to quickly fix the focus bug that you filed
16:36:50 <joanie> but not in a position to fix it himself or with red hat devs
16:36:51 <API> but afaik, they are planning to switch gtkthml wth webkitgtk, right?
16:36:54 <mgorse> but maybe not
16:37:12 <joanie> that's my understanding API
16:37:30 <mgorse> I thought that was the plan. They've switched some things so far but not others
16:37:44 <jjmarin> AFAIK, aruiz is managing the development of evolution
16:37:54 <mgorse> I also have some patches that I need to file a new bug for
16:37:55 <joanie> oh
16:37:58 <joanie> hmmmm
16:38:03 <joanie> maybe it's aruiz
16:38:17 <joanie> anyhoo, we should find out and ping them until they submit
16:38:24 <API> well, he was the one that sent that big blog post "Evolution moving to one-year cycle"
16:38:32 <joanie> my bad
16:38:54 <mgorse> mbarnes is one of the main people, and I get the impression he just would like to have someone else maintain the a11y code, going by a comment he made a while ago when removing some code that was crashing
16:39:00 <mgorse> probably he is busy in general
16:39:20 <joanie> would it make sense for someone (either API or mgorse) to ping aruiz?
16:39:45 <jjmarin> I think aruiz is the right person
16:40:03 <joanie> because with the Thunderbird a11y bugs and the Evo a11y bugs, we have a situation where "Orca doesn't provide access to email"
16:40:17 <joanie> and we need *some* properly accessible gui-based email client
16:40:23 <mgorse> I could ping him. I'd like to get the message composer issues resolved anyhow
16:40:25 <joanie> that we lack this is embarrassing
16:40:38 <joanie> mgorse: please take an action item for that then?
16:40:57 <mgorse> #action mgorse will ping aruiz re: evolution and gtkhtml accessibility bugs
16:41:02 <joanie> thanks!!
16:41:11 <joanie> mgorse: related to the bugs you plan to file, could you cc me?
16:41:19 <joanie> and use keyword accessibility
16:41:37 <mgorse> okay. Afaik there's no "a11y hasn't been refactored for gtk+ 3.2" bug yet
16:41:48 <mgorse> or similar, except the ones concerning gtkhtml
16:41:53 <joanie> k
16:42:02 <API> ok, so jjmarin
16:42:05 <API> now is the moment ;)
16:42:08 <joanie> heh
16:42:14 * joanie re-re-shutsup
16:42:14 <jjmarin> :-)
16:42:17 <jjmarin> #info The pointer size is not configurable in GNOME 3. The main bug is #608231 where Allan Day asks the opinion of the a11y party. There is a person, Jason Steward, who volunteers to code the fix and is waiting for some confirmation.
16:42:29 <jjmarin> #info BTW, Allan Day has some open questions in (duplicated) #665907 about the relation to the size with the HC theme then and about the color effects and the pointer.
16:43:00 <jjmarin> that's all I think about this particular topic
16:43:08 <API> HC theme?
16:43:13 <API> ah high contrast
16:43:17 <jjmarin> High Contrast
16:43:28 <API> hmm, I don't see any relation between high contrast theme and pointer size
16:43:35 <clown> bug 608231
16:43:35 <tota11y> 04Bug https://bugzilla.gnome.org/show_bug.cgi?id=608231 enhancement, Normal, ---, control-center-maint, UNCONFIRMED, Large Cursor For Low Vision
16:43:46 <API> I mean that both should be configurable independently
16:43:58 <API> imho
16:44:02 <joanie> but high contrast theme should change the pointer color potentially
16:44:12 <API> ah
16:44:17 <API> yes, that's true
16:44:27 <jjmarin> I guess, mclassen also want to find some relation with text size
16:44:28 <joanie> black pointer on a navy entry is teh suck :)
16:44:36 <jjmarin> pendulum said it was not a good idea
16:44:37 <API> just focused on "Allan Day has some open questions in (duplicated) #665907 about the relation to the size with the HC theme"
16:44:44 <API> so for that: no relation
16:44:51 <API> " and about the color effects and the pointer."
16:44:56 <API> for that: yes, it shoudl change too
16:45:45 <API> in any case, anyone knows Jason Stewart?
16:45:53 <jjmarin> I don;t
16:46:04 <clown> didn't the old gnome have something like this:  different mouse cursor sizes?
16:46:11 <API> is somewhat odd that he is saying that will solve the bug if someone confirms that is a bug
16:46:23 <API> if he is not able to reproduce the bug, how will solve it?
16:46:54 <API> clown, yes I think so, was somewhat missed with the switch
16:46:56 <joanie> i think perhaps he means, if someone agrees that we have a problem
16:46:57 <jjmarin> I guess it must be confirmed it is bug, not a evil design decision
16:47:06 <joanie> exactly
16:47:21 <joanie> mouse pointers are distracting
16:47:24 <API> well, I don't think that was removed by purpose as a design decision
16:47:35 <API> in any case
16:47:41 <API> jjmarin, as you bring this here
16:47:54 <API> can we assign you to track this bug?
16:48:05 <API> I also think that you made some research some time ago
16:48:06 <jjmarin> ok
16:48:16 <API> so something like "confirming" is a bug, and so on
16:48:25 <API> after all, we already have someone volunteering to solve it
16:48:34 <API> it would be good to know if he will be able to do it or not
16:48:37 <jjmarin> but my question before of this, where the pointer configuration must be
16:48:49 <API> well, thats is a design
16:48:59 <API> so Allan Day or any of his colleagues/minions
16:49:02 <API> decision
16:49:08 <jjmarin> in a11y/Seeing I guess
16:49:34 <API> so, as he was asking question to the "a11y people", could you provide answers and follow it?
16:49:43 <API> you could also suggest places, of course
16:49:50 <API> probably that could be a starting point
16:50:05 <jjmarin> #action Juanjo will track bug 608231
16:50:05 <tota11y> 04Bug https://bugzilla.gnome.org/show_bug.cgi?id=608231 enhancement, Normal, ---, control-center-maint, UNCONFIRMED, Large Cursor For Low Vision
16:50:09 <API> jjmarin, ok thanks
16:50:13 <API> and as we don't have too much time
16:50:16 <API> lets move
16:50:17 <jjmarin> ok, thanks
16:50:24 <API> #topic W·C updates
16:50:25 <API> ups
16:50:30 <joanie> hahaha
16:50:31 <API> #topic W3C updates
16:50:38 <API> clown, joanie ?
16:50:41 <jjmarin> Lol
16:50:46 * joanie defers to clown
16:50:57 * clown collects his thoughts.
16:51:20 <clown> #info Reminder that there is a three day face-to-face ARIA meeting here in Toronto next week.
16:51:39 <clown> #info Thu Jan 23 - Sat Jan 25
16:52:14 <clown> #info This week is the end of the "any last comments on the UAIG" before moving to the next step for publication.
16:52:21 <joanie> congrats
16:52:42 <clown> #info  As of Mon, the four features at risk will be considered and dealt with:  http://www.w3.org/TR/wai-aria-implementation/#sotd_atrisk
16:53:13 <clown> #info Summarizing the probable outcomes:  Feature one will be removed as there are no implementations,
16:53:39 <clown> #info Feature two will likely be removed, unless Microsoft tells us otherwise.
16:53:51 <clown> #info Feature three will stay — there are implemenations.
16:54:20 <clown> #info Feature four:  we are waiting on Apple to confirm that they implemented it.  It will likely stay.
16:54:42 <clown> done questions?  (and thanks, joanie).
16:54:53 <joanie> clown: fwiw I keep seeing bugzilla spam about it for webkit
16:55:02 <API> well, without knowing too much about it
16:55:11 <API> just reading feature 1 summary
16:55:28 <API> "keyboard accessibility by event simulation" sounds somewhat
16:55:36 <API> strange to be something to be normative
16:55:48 <API> but as I said, I don't know the details of that feature
16:56:04 <clown> Well, it's talking about 10 and 11.  Ten is:
16:56:42 <clown> "When the user triggers an element that is only focusable because of its tabindex attribute in a manner other than clicking it, such as by pressing Enter, and the element has no defined activation behavior fire a click event."
16:57:17 <clown> or, more briefly, if the user activate a control via ENTER, simiulate a click event.
16:57:29 <joanie> which makes good sense to me
16:57:43 <clown> but no one implements it for ARIA widgest.
16:57:51 <clown> *widgets.
16:58:03 <clown> native widgets like <button> show the behaviour.
16:58:05 <magpie> what about hover?
16:58:21 <magpie> does that not focus?
16:58:26 <clown> no.
16:58:30 <clown> not generally, anyway.
16:58:31 <magpie> hmm
16:58:33 <joanie> hover hovers
16:58:48 <clown> sometimes it "arms", but it does not move focus.
16:58:59 <API> in any case, we don¡t have too much time
16:59:13 <API> so if you don't mind, I will move to next topic
16:59:14 <clown> just one more comment.
16:59:19 <API> k
16:59:30 <clown> to finish out the point raised by joanie
16:59:39 <clown> That is should work that way.
17:00:03 <clown> This is one of the features to be re-introduced in ARIA 1.1, and will be addressed somewhat at next week's meeting.
17:00:04 <clown> done.
17:00:08 * joanie nods
17:00:10 <joanie> thanks
17:00:15 <clown> no problem.
17:01:03 <API> so moving
17:01:06 <API> #topic Marketing
17:01:17 <API> a lot of weeks without marketing update, I blame christmas
17:01:20 <API> jjmarin, something to say?
17:01:30 <jjmarin> sorry, nothing so far
17:01:39 <jjmarin> I promise work this week :-)
17:01:42 <API> ok, then this was easy
17:01:43 <joanie> :)
17:01:51 <API> np, we already assigned you a task today
17:01:54 <API> #topic misc time
17:02:08 <API> #info FOSDEM is coming
17:02:18 <jjmarin> API, I sent you an email about the Spanish podcast
17:02:25 <API> #info there will be GNOME activity there
17:02:37 <magpie> i might go to that
17:02:40 <API> #info this year API will not be there to give the usual a11y related presentation
17:03:14 <API> #info but is possible one co-worker will be there, and attend the after-fosdem libreoffice hackfest, probably doing a11y stuff
17:03:21 <API> jjmarin, ok,
17:03:32 <API> true, we talked about that some meetings ago
17:03:40 <API> I will read the email after meeting
17:03:40 <API> thanks
17:03:45 <API> so that was my misc stuff
17:03:57 <API> anyone else want to share something misc (and quick)?
17:04:15 <jjmarin> just to make sure it wasn't killed by the anti spam measures :-)
17:04:30 <magpie> yeah we've only mentioned physical disabilities in the landing pages
17:05:09 <jjmarin> magpie: what do you mean with landing pages ?
17:05:42 <magpie> https://help.gnome.org/users/gnome-help/stable/a11y.html
17:06:15 <magpie> https://wiki.gnome.org/Accessibility/
17:06:54 <jjmarin> many of these physical disabilities can be derived for cognitive disabilities
17:07:00 <jjmarin> from
17:07:07 <magpie> there's no mention of cognitive or seisure disorders
17:07:39 <magpie> and gnome has not accessibility statement on how to use the site
17:08:06 <jjmarin> in fact, this is the online version of the documentation from yelp
17:08:27 <magpie> so this is documentation bug?
17:09:12 <API> ok, seems a good point to be raised here
17:09:13 <API> *but
17:09:20 <API> a point to be elaborated outside
17:09:25 <jjmarin> ok
17:09:30 <magpie> ok
17:09:31 <API> so if you don't mind, I will close the meeting
17:09:43 <API> and the details could be elaborated somehwere else
17:09:48 <jjmarin> go ahead !!!
17:09:48 <magpie> a list of current bugs would be good if there isn;t already
17:09:57 <API> so closing the (over-time) meeting
17:10:12 <API> and if there aren't opening it
17:10:19 <API> #endmeeting