[aapi] Minutes: UAI TF Meeting Tue Jan 19, 2016

Link: https://www.w3.org/2016/01/19-aapi-minutes.html

Plain text follows.
   [1]W3C

      [1] http://www.w3.org/

   Accessible Rich Internet Applications Working Group Teleconference

19 Jan 2016

   See also: [2]IRC log

      [2] http://www.w3.org/2016/01/19-aapi-irc

Attendees

   Present
          Joseph_Scheuhammer, Joanmarie_Diggs, Bryan_Garaventa,
          Rich_Schwerdtfeger, Cynthia_Shelly

   Regrets
   Chair
          Joseph_Scheuhammer

   Scribe
          joanie, Rich, Joseph

Contents

     * [3]Topics
         1. [4]ACTION-1531 (All) Bolter to investigate ia2 mapping
            of role="text"
         2. [5]ACTION-1533 (Joanie) Possibly create minimal
            JavaScript AT-SPI listener for automating CR tests.
         3. [6]ACTION-1535 (Joseph/James) AXAPI mappings for
            aria-current (notifications).
         4. [7]ACTION-1547 (Rich) Update test cases involving
            rowgroup.
     * [8]Summary of Action Items
     * [9]Summary of Resolutions
     __________________________________________________________

   <clown> agenda: this

   <joanie> scribe: joanie

ACTION-1531 (All) Bolter to investigate ia2 mapping of role="text"

   <clown> action-1531?

   <trackbot> action-1531 -- David Bolter to Bolter to investigate
   ia2 mapping of role="text" -- due 2014-11-18 -- OPEN

   <trackbot> [10]http://www.w3.org/WAI/ARIA/track/actions/1531

     [10] http://www.w3.org/WAI/ARIA/track/actions/1531

   JS: I want to follow up on something discussed last week.
   ... It's about the IA2 mapping for role="text"
   ... I mentioned a mailing list discussion when the agreement
   was IA2_ROLE_TEXTFRAME

   <clown>
   [11]https://lists.w3.org/Archives/Public/wai-xtech/2015Sep/0017
   .html

     [11] https://lists.w3.org/Archives/Public/wai-xtech/2015Sep/0017.html

   JS: I found the email which captures it. See URL above.
   ... Background: Rich wrote to this list asking Alex and James
   Teh what their expectation was with respect to the textframe
   role and the text interface.
   ... (reads from thread cited above)
   ... At the end, James Teh asked if ROLE_SYSTEM_STATIC_TEXT was
   concerned.
   ... I don't know what that means, and no one answered him.

   RS: It's always been for things like text displayed in message
   boxes.
   ... So the idea is you're going to have role="text" and then a
   label, right?

   JS: Not necessarily.

   <clown>
   [12]http://rawgit.com/w3c/aria/master/core-aam/core-aam.html#ro
   le-map-text

     [12]
http://rawgit.com/w3c/aria/master/core-aam/core-aam.html#role-map-text

   RS: Let's say you put role="presentation" on a table.

   <richardschwerdtfeger> <td> Hello Joanie <img
   src=“collegephoto.jpg” alt=“crazywoman”> Diggs</td>

   RS: What would you do with the above?

   <clown>
   [13]http://rawgit.com/w3c/aria/master/core-aam/core-aam.html#ro
   le-map-presentation

     [13]
http://rawgit.com/w3c/aria/master/core-aam/core-aam.html#role-map-presentation

   Group: The TD is a textframe in IA2.

   RS: I think the same thing will happen with role="text"

   <richardschwerdtfeger> scribe: Rich

   <richardschwerdtfeger> scribe: Joseph

   <clown> RS: Let say we have a row of these cells with content
   inside of them, how would they be read?

   <clown> JD: They're not mapped as text frames on my platform,
   but as separated block elements.

   <clown> JD: But with text frames, I'd have to look on a windows
   machine

   <clown> RS: I think you want to pause between the pieces.

   <clown> RS: They might come out as groups in AXAPI. And also
   UIA?

   <clown> CS: Not on UIA. They are just text elements.

   <clown> CS: I'd have to look to be sure.

   <clown> CS: To be sure if there are separate text chunks, or as
   a one long string.

   <clown> JD: I heard what you said and see the core-aam, but I
   question that FF is doing that.

   <clown> JD: Surkov made it such that ATK would map to
   ROLE_STATIC.

   <clown> JD: At my request, Surkov now maps span-like elements
   that have a property that requires an accessible object are
   being mapped as ATK_ROLE_STATIC

   <clown> JD: For all of those instances (<abbr>, <span> with
   onclick) are mapped on IA2 as text frames.

   <clown> JD: Ergo, IA2_ROLE_TEXTFRAMES are treated as inlines.

   <clown> JD: This is what I think is happening.

   <clown> RS: Confirming that they become text frames in IA2.

   <clown> CS: In UIA they are no longer in the a11y tree. They
   are just part of a text pattern.

   <clown> CS: Even so, narrator is treating them as paragraphs
   (the cells).

   <clown> RS: So they are presented as separate.

   <clown> RS: What about Linux?

   <clown> JD: Each cell is mapped as a DIV. The role is
   ROLE_SECTION.

   <clown> RS: In all cases, then, they are separate blocks.

   <clown> RS: the problem with text frame is it is seen as a
   container.

   <clown> JD: For role="text".

   <clown> RS: Yes.

   <clown> RS: If you had a text frame followed by text, how would
   you read it?

   <clown> JD: the text would be inline.

   <clown> RS: If you want it inline, then you don't want a text
   frame.

   <clown> RS: We want the same experience on all platforms.

   <clown> RS: UIA would map it as a text pattern, right Cynthia.

   <clown> CS: Yes. Text pattern has a way of marking spaces and
   format changes. That's what narrator is keying off of.

   <clown> RS: In the case of role text, we want it to be read
   together.

   <clown> RS: Would that happen onf UIA?

   <clown> CS: Yes

   <scribe> scribe: joanie

   <clown>
   [14]https://lists.w3.org/Archives/Public/wai-xtech/2015Sep/0017
   .html

     [14] https://lists.w3.org/Archives/Public/wai-xtech/2015Sep/0017.html

   JD: So I think the next step is for someone to respond to that
   thread and say that we think James is right re
   ROLE_SYSTEM_STATICTEXT.

   JS: I'd like Rich to do that.

   RS: I'm going to respond to xtech and refer to the thread.

   <clown> wai-xtech@w3.org

   <clown> IA2_ROLE_TEXTFRAME

   RS: I'll also CC public-aria
   ... Unless they say something different, this should be done.

   JS: So let's see what they say.

ACTION-1533 (Joanie) Possibly create minimal JavaScript AT-SPI
listener for automating CR tests.

   <clown> action-1533?

   <trackbot> action-1533 -- Joanmarie Diggs to Investigate and
   possibly create a minimal/skeleton javascript at-spi2 listener
   with the aim of automating the 1.1 tests. -- due 2016-02-23 --
   OPEN

   <trackbot> [15]http://www.w3.org/WAI/ARIA/track/actions/1533

     [15] http://www.w3.org/WAI/ARIA/track/actions/1533

   JD: I already postponed it until the end of February.

ACTION-1535 (Joseph/James) AXAPI mappings for aria-current
(notifications).

   <clown> action-1535?

   <trackbot> action-1535 -- James Craig to Investigate the axapi
   mappings for aria-current. -- due 2015-08-18 -- OPEN

   <trackbot> [16]http://www.w3.org/WAI/ARIA/track/actions/1535

     [16] http://www.w3.org/WAI/ARIA/track/actions/1535

   JS: This is more of an information update.
   ... Everything is done for mapping aria-current, except for one
   thing.
   ... Will Apple emit a notification?

   <clown> [17]https://bugs.webkit.org/show_bug.cgi?id=146012#c19

     [17] https://bugs.webkit.org/show_bug.cgi?id=146012#c19

   JS: The answer is at the above URL.

   <clown> AXAriaAttributeChanged

   JS: There's no specific notification for aria-current, but
   there will be for the attribute change.
   ... I imagine it's emitted any time any ARIA attribute is
   changed.
   ... I replied to that saying that I assume that is a general
   notification and asking if they want that in the mapping spec.

   RS: Are state changes listed?

   JS: Yes they are in there.

   RS: What about ARIA attributes?

   JS: No.

   RS: Then we probably need one.

   JS: Do we?
   ... (Reads example of other attributes for AX API listed in the
   mapping spec)

   RS: What other attributes besides aria-current will trigger
   this event?

   JS: All of them.

   RS: Oh. So we definitely need to say something.

   JS: I believe AXAriaAttributeChanged is internal; not to be
   exposed to ATs. But I don't know.

   RS: If we can't test it....

   JS: If you put this AXAriaAttributeChanged in the table, you're
   telling ATs that they can listen for that.

   RS: But how else would you do it?

   JS: This is the only way, but it's not performant.

   RS: What I think you want to do is list those things you think
   you should be watching for.
   ... But if we put it in the table, you have to be able to test
   it.
   ... Does the inspect tool let you listen for events?

   JS: This is the case in Windows and Linux, but I don't know on
   the Mac.
   ... I asked in Bugzilla if they really want me to put this
   event in the mapping table for aria-current notifications.
   ... If they say yes, then I'll ask them how we can test it.
   ... So that's the update.

ACTION-1547 (Rich) Update test cases involving rowgroup.

   <clown> action-1547

   <trackbot> action-1547 -- Richard Schwerdtfeger to Update test
   test cases involving rowgroup to ensure they are included
   within a grid container and to ensure that the expected result
   matches the new mappings -- due 2015-10-27 -- OPEN

   <trackbot> [18]http://www.w3.org/WAI/ARIA/track/actions/1547

     [18] http://www.w3.org/WAI/ARIA/track/actions/1547

   JS: This action is for you (Rich) to update test cases
   involving rowgroup.

   RS: Should this be in the test harness?
   ... If we move it to the ARIA 1.1 product, we don't have to
   discuss it here.

   JS: I'll do that.

   issue-635

   <trackbot> issue-635 -- Determine if uaig mappings for rowgroup
   are correct -- open

   <trackbot> [19]http://www.w3.org/WAI/ARIA/track/issues/635

     [19] http://www.w3.org/WAI/ARIA/track/issues/635

   JS: It's associated with issue-635, which is the core mapping
   specification.
   ... We'd have to decouple it from this issue.

   RS: Have we got the rowgroup mappings determined?

   JS: Up to a point.

   <clown>
   [20]http://rawgit.com/w3c/aria/master/core-aam/core-aam.html#ro
   le-map-rowgroup

     [20]
http://rawgit.com/w3c/aria/master/core-aam/core-aam.html#role-map-rowgroup

   RS: We could write the test cases if we have the mappings.

   JS: Ok. I'll associate it with the test harness.
   ... I put the URL in for the mappings (above)
   ... (Reads them)
   ... AX API is not mapped.

   JD: I remember a bunch of discussion, but was that the
   definitive conclusion?

   <clown> [21]https://bugs.webkit.org/show_bug.cgi?id=146239#c6

     [21] https://bugs.webkit.org/show_bug.cgi?id=146239#c6

   JS: James replied that if the focusable rowgroup must be
   exposed, then he would not expose the table as a table because
   it's an authoring error.

   JD: Perhaps Steve Faulkner could use his tool to address James'
   question about real-world use cases

   JS: I'll contact Steve about this.

   <clown> action-1686

   <trackbot> action-1686 -- Joseph Scheuhammer to Drive mapping
   of aria-errormessage to aapis -- due 2015-07-23 -- OPEN

   <trackbot> [22]http://www.w3.org/WAI/ARIA/track/actions/1686

     [22] http://www.w3.org/WAI/ARIA/track/actions/1686

   <clown>
   [23]http://rawgit.com/w3c/aria/master/aria/aria.html#aria-error
   message

     [23] http://rawgit.com/w3c/aria/master/aria/aria.html#aria-errormessage

   <clown>
   [24]http://rawgit.com/w3c/aria/master/core-aam/core-aam.html#ar
   iaDescribedBy

     [24]
http://rawgit.com/w3c/aria/master/core-aam/core-aam.html#ariaDescribedBy

   <richardschwerdtfeger>
   [25]https://rawgit.com/w3c/aria/master/aria/aria.html#aria-erro
   rmessage

     [25]
https://rawgit.com/w3c/aria/master/aria/aria.html#aria-errormessage

Summary of Action Items

Summary of Resolutions

   [End of minutes]

Received on Tuesday, 19 January 2016 21:35:14 UTC