14:05:34 <API> #startmeeting
14:05:35 <Services> Meeting started Thu Oct  9 14:05:34 2014 UTC.  The chair is API. Information about MeetBot at http://wiki.debian.org/MeetBot.
14:05:35 <Services> Useful Commands: #action #agreed #help #info #idea #link #topic.
14:05:47 <API> #topic Progress towards 3.14
14:05:55 <API> (I hope that last one on this cycle)
14:06:01 <joanie> heh
14:06:07 <API> #topic Next monday 13  is the release of 3.14.1
14:06:09 <API> ah
14:06:12 <API> #topic Progress towards 3.14
14:06:17 <API> #info Next monday 13  is the release of 3.14.1
14:06:57 <API> #info Main fix to be included would be the one to avoid event floods, from bug 728319
14:07:08 <Services> 04Bug http://bugzilla.gnome.org/show_bug.cgi?id=728319 normal, Normal, ---, at-spi-maint, RESOLVED FIXED, Hangs when browsing using artist view
14:07:08 <API> https://bug728319.bugzilla-attachments.gnome.org/attachment.cgi?id=286996
14:07:13 <API> mgorse, so this is a kind of reminder
14:07:17 <API> and done
14:07:30 <API> so if anyone want to coment other new stuff for 3.14.1
14:07:39 <joanie> not me
14:08:11 <mgorse> not me; my recent commits are only going into 3.15
14:08:41 <mgorse> and thanks for the reminder. Made a note to myself to make a release
14:09:58 <API> ok, then, this point will be short
14:10:15 <API> something that is good as we are talking about 3.14.0->3.14.1 updates
14:10:41 <API> probably next week we should talk about big things (if needed) to focus on for the next cycle
14:10:48 <API> well, next cycle that in fact is the current cycle
14:10:52 <API> we are already on 3.15
14:10:53 * joanie nods
14:10:57 * clown waves
14:11:05 <API> so just as a reminder
14:11:43 <API> #info Next week we will start the meeting with a "3.15 is starting", and we will try to list the main objectives for the cycle
14:12:36 <API> #info Just as an idea a) Spread the past new APIs on the toolkits (ie: ATK new text API are still not implemented in several toolkits)
14:13:05 <API> #info b) New APIs (there are plans for several new APIs on ATK/AT-SPI, probably we should be conservative and target for one or two)
14:13:07 <API> an done
14:13:09 <API> so
14:13:11 <API> next topic=
14:13:12 <API> ?
14:14:26 <API> seems so
14:14:29 <API> #topic W3C updates
14:14:34 <API> joanie, clown ?
14:14:43 * joanie defers to clown as always
14:14:57 <clown> I have one (difficult to explain quickly).
14:15:48 <clown> #info  A feature of CSS has come up that some in the working group are having issues with.
14:16:05 <clown> #info CSS has the ability to add content from within the style sheet.
14:16:54 <clown> #info When calculating the accessible name from the markup, browsers need to be aware of this, and include CSS content if the CSS is used in this way (using content rules).
14:17:14 <clown> #info Here is an example:  http://clown.idrc.ocad.ca/Fluid/aria/renderedText.html
14:17:43 <clown> #info The text "We should" and "it rains" in the first sentence is not in the DOM.  It's coming from the CSS.
14:17:59 <joanie> wheeee
14:18:08 <joanie> thanks for this test case btw
14:18:10 <clown> #info  As I said, some in the working group find this problematic.  In fact, you can't even search for that text in FF.
14:18:44 <clown> #info Try to do a select all.  Only the text in the DOM is highlighted, not the content coming from the CSS.
14:19:40 <clown> #info  The face-to-face meeting at the end of the month at TPAC:  the PF group will try to meet with the CSS group and discuss this.
14:19:48 <clown> joanie, did I miss anything?
14:19:58 <joanie> not that I recall
14:20:09 <joanie> I do want to see what is exposed by both Gecko and webkit
14:20:11 <clown> okay, done for me, then.  questions?
14:20:20 <API> although Im not a designer
14:20:24 <API> as far as I understood
14:20:34 <API> CSS is about appeareance and all that jazz
14:20:46 <API> so what is the purpose of this feature allowing CSS to add content?
14:20:47 <clown> I can check quickly, joanie, with respect to FF's own internal a11y tree.
14:21:17 <clown> API, that's the question I always ask when this comes up.  I don't remember the answer.
14:21:31 <joanie> others have asked it too
14:21:46 <joanie> and I don't think we've gotten an answer. Maybe that's for TPAC?
14:22:00 <API> so I assume that part of this "lets talk with CSS group" is getting an answer to that question?
14:22:38 <clown> well, sometimes the answer is;  That ship has sailed.  The CSS content rules have been in place for many years now.
14:22:56 <joanie> oh, true, I do now remember hearing that answer
14:23:22 <joanie> along with a "we really need all of these groups to communicate and work together more effectively/closely"
14:23:24 <API> so the answer is "that doesn't matter because is being there for years"?
14:23:40 <joanie> not "doesn't matter"
14:24:40 <clown> more like, too many web pages already use it, so we have to deal with it.  "We" = accessibility software.
14:24:52 <joanie> clown: I see the text in order in accerciser
14:24:53 <joanie> yay!
14:25:01 <joanie> for FF anyway
14:25:06 * joanie tries epiphany
14:25:10 <API> hmm, ok, I was just guessing if CSS can just remove that feature or something
14:25:11 <clown> yes, FF does a reasonable job of handling this CSS content.
14:25:17 <API> remove as deprecate
14:25:23 <API> in any case, no more questions from my side
14:25:32 <clown> btw, FF"s a11y tree shows three paragraph accessibles.  The CSS content is listed separately as "static text", while the DOM text is listed as "text leaf".
14:25:42 <jjmarin> good question I think :-)
14:26:01 <joanie> i'm happy with paragraphs
14:26:17 <joanie> it looks paragraphy
14:26:19 <joanie> ;)
14:26:30 <joanie> not sure what this means for caret nav
14:26:36 <clown> I think one of the issues for TPAC is that CSS wants to extend the content rules.  PF will try to argue against that.
14:26:40 <joanie> not navigable
14:26:44 <joanie> wcag violation
14:26:59 <clown> well, as I said you can search the text in the browser even, never mind the a11y layer.
14:27:06 <clown> whoops.
14:27:13 <clown> you *cannot* search the text.
14:27:16 <joanie> i think the a11y layer has more access
14:27:32 <joanie> sighted users are on their own :P
14:27:51 <clown> do a ctl-f and search for "We should".  FF will say, "text not found".
14:29:26 <jjmarin> New way to send crypto messages :-)
14:29:43 <clown> good point, jjmarin
14:29:58 <clown> a way to hide text from search engines.
14:30:16 <joanie> webkit is not exposing it
14:30:17 <joanie> boo
14:30:46 <joanie> #action Joanie will file a bug against WebKitGtk (or WebKit, pending checking on her Mac) about this test case.
14:31:02 <clown> I thought James C said that Safari supported it.  Or is safari's verions of webkit ahead of webkitgtk?
14:31:12 <joanie> it's not "ahead"
14:31:19 <clown> behind?
14:31:21 <clown> ;-)
14:31:23 <joanie> it's that there is WebCore accessibility
14:31:38 <joanie> and then there's the platform a11y
14:31:54 <joanie> if Safari exposes it correctly, then it's in WebCore
14:32:03 <joanie> and maybe further in the Mac
14:32:09 <joanie> further == additional support
14:32:13 <joanie> though not necessarily
14:32:27 <joanie> in which case the ATK implementation needs this added
14:32:40 <joanie> but first I want to verify this in a nightly build of Safari
14:32:49 <joanie> anyhoo, this is a deep dive now
14:32:51 <clown> sure.
14:33:15 <clown> I'm done.  Anything that joanie wants to add to this topic, well, I hand her the conch.
14:33:21 <API> ok
14:33:23 <API> so moving then?
14:33:25 <joanie> nah, I'm done.
14:33:29 <joanie> yeah re moving
14:33:47 <API> #topic Marketing
14:33:50 <API> jjmarin,
14:33:50 <API> ?
14:34:01 <jjmarin> not too much this week.
14:34:08 <jjmarin> #info By now, the idea of talking about the Debian adoption wasn't very good received. Mainly because it is not clear if the decision is already taken or not, and because I wasn't able to talk to the much people about it on the IRC.
14:34:18 <jjmarin> I'll see if I can do more about this
14:34:20 <jjmarin> done !
14:34:32 <jjmarin> :-)
14:34:37 <joanie> I tried to write the promised blog post
14:34:47 <joanie> but it led to 20 years of history
14:34:51 <joanie> and I set it aside
14:35:21 <jjmarin> sounds like you're writing a book of memories :-)
14:35:32 <clown> autobiography?
14:35:38 <joanie> no
14:35:49 <joanie> it's a long story
14:35:52 <joanie> short version:
14:36:28 <joanie> the JAWS virtual buffer has subsequently dictated what web devs and users expect wrt a11y from screen readers
14:36:48 <joanie> it's why I had to rewrite much of Orca's support for Gecko
14:37:04 <joanie> and it's why I'm going to have to implement an Orca-controlled caret for Epiphany/WebKit
14:37:30 <joanie> and I also want to explain how Orca doesn't have a virtual buffer
14:37:36 <joanie> and accomplishes what it does
14:37:53 <joanie> there are also assumptions from the early days of ARIA that are no longer true
14:38:00 <joanie> namely, ARIA is for "web apps"
14:38:17 <joanie> that, too, required much of the Great Rewrite of Orca's gecko support
14:38:19 <joanie> anyhoo....
14:38:36 <jjmarin> I buy your story, sounds interesting and informative. And you're good writing :-)
14:38:48 <joanie> this has been a hard one to write
14:38:48 <clown> I think you just wrote the gist of your blog.  Give it jjmarin to flesh out :-)
14:38:54 <joanie> heh
14:39:30 <jjmarin> yes, you can pass me what you have if you want
14:39:50 <joanie> there's more info needed
14:39:53 <joanie> which is in my head
14:39:55 <joanie> not on paper
14:40:15 <jjmarin> ok, good plot so far :-)
14:40:27 <clown> someone should invent a device that transfers things in one's head immediately to paper.
14:40:36 <clown> I would find it really useful...
14:40:47 <joanie> I would as well
14:40:53 <API> just asking, but until now JAWS dictated what users expect from a screen reader, didn't all the mobile screen reader (VoiceOver, Talkback) changed that dictation somehow?
14:41:12 <joanie> no :P
14:41:45 <joanie> i mean, they are instances that are virtualbufferless
14:41:47 <joanie> afaik
14:42:08 <API> well, yes,
14:42:11 <joanie> but the expecation that a screen reader presents -- and allows the user to navigate within -- content based on the a11y tree
14:42:24 <joanie> rather than a screen reader presents navigation in the actual content
14:42:28 <joanie> is already set
14:42:44 <joanie> so if Orca just presents what happens as the user navigates in Epiphany
14:42:55 <joanie> web devs say "Orca doesn't support x and y"
14:43:24 <joanie> because x and y is not something you can interact with/encounter as a sighted user navigating the document content
14:43:34 <API> I mean that in the past, you needed to justify constantly why Orca doesn't have virtual buffer, and a lot of time (afair), the answer was "all relevent screen readers uses virtual buffer", at least now there are example of other buffer-less screen readers
14:43:43 <joanie> yes
14:43:52 <joanie> but they navigate to things based on the a11y tree
14:43:55 <joanie> not the render tree
14:44:05 <joanie> example: off-screen elements
14:44:19 <joanie> so you can either put stuff in a virtual buffer
14:44:27 <joanie> or you can just navigate as you go
14:44:37 <joanie> Orca, and I believe VoiceOver, do the latter
14:44:45 <joanie> well, Orca does that for Gecko
14:44:51 <joanie> not (yet) epiphany
14:45:21 <joanie> in other words, the virtual buffer is now an implementation detail
14:45:42 <joanie> it is one way to make it possible for users who are blind to navigate in what web devs make available
14:46:09 <joanie> the expectation that the navigable content is more than just what's rendered on screen, however, is set
14:46:19 <joanie> too late to do anything about that
14:46:37 <joanie> ATs just need to figure out the best way for their platform to make that navigation possible
14:46:43 <joanie> you follow?
14:46:50 <API> yes, thanks for the explanation
14:47:48 <API> and I will wait for that mythical post for the extra details
14:47:54 <joanie> heh
14:47:56 <API> perhaps this is too detailed for the meeting
14:47:57 <API> so
14:48:03 <API> more questions on this topic¿
14:48:04 <API> moving?
14:48:13 <joanie> you are the one who asked :P
14:48:36 <jjmarin> :-)
14:49:42 <clown> *crickets*
14:49:49 <API> #topic Miscellaneous time
14:49:57 <API> something not included on agenda to mention?
14:50:09 <clown> when does summer time end in Europe/Spain?
14:50:22 <clown> Our daylight savings time changes Nov 2.
14:51:11 <jjmarin> no idea here, API ?
14:51:34 <API> probably on the 25-25 weekend
14:51:47 <API> but not sure
14:52:20 <jjmarin> meanwhile, we can start the kickstart campaign for the head-to-paper device that clown and joanie need :-)
14:52:38 <clown> okay, probably only one meeting on Oct 30 that I will have to remind myself of the time change.
14:53:13 <clown> Maybe there is no meeting on Oct 30 for joanie and I since I think that is when TPAC is.
14:57:55 <API> http://www.timeanddate.com/time/change/spain/madrid
14:58:02 <API> it seems that the change is on 30 october
14:58:05 <API> sorry
14:58:08 <API> 26 october
14:58:26 <clown> thanks API
14:58:31 <API> so, as the question is answered, I think that we can safely close the meeting
14:58:35 <API> thanks everybody for coming
14:58:41 <API> #endmeeting