[AAPI] Minutes UIA TF Meeting Tuesday, 31 January 2017

Link: https://www.w3.org/2017/01/31-aapi-minutes.html

Plain text follows:

   [1]W3C

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

   Accessible Rich Internet Applications Working Group Teleconference

31 Jan 2017

   See also: [2]IRC log

      [2] http://www.w3.org/2017/01/31-aapi-irc

Attendees

   Present
          Joanmarie_Diggs, Joseph_Scheuhammer, Rich_Schwerdtfeger,
          Bryan_Garaventa

   Regrets
   Chair
          Joseph_Scheuhammer

   Scribe
          joanie

Contents

     * [3]Topics
         1. [4]GH-ISSUE-513: (All/Joanie) Mapping of role="region"
            when it doesn't have an accessible name.
         2. [5]ACTION-1707 (Joseph/All) pending final review.
         3. [6](All) Specifying AAPI UIA interface in focus events
            table
         4. [7](All) update on AX API GH-ISSUEs
         5. [8](All) aria-label and the accname-aam
         6. [9](All) aria-labelledby and the accname-aam
         7. [10]Triage issues/actions:
            https://www.w3.org/WAI/ARIA/track/products/23
     * [11]Summary of Action Items
     * [12]Summary of Resolutions
     __________________________________________________________

   <clown> agenda: this

   <scribe> scribe: joanie

GH-ISSUE-513: (All/Joanie) Mapping of role="region" when it doesn't
have an accessible name.

   <clown>
   [13]https://github.com/w3c/aria/issues/513#issuecomment-2761502
   66

     [13] https://github.com/w3c/aria/issues/513#issuecomment-276150266

   JS: This is the recent request to change the mapping for
   role="region" if it doesn't have a name.

   <clown>
   [14]https://rawgit.com/w3c/aria/master/core-aam/core-aam.html#r
   ole-map-region

     [14]
https://rawgit.com/w3c/aria/master/core-aam/core-aam.html#role-map-region

   JS: We agreed to the changes, which I made (see URL above)
   ... For each AAPI, I added a special case to use the native
   host language mapping rather than landmark when the region has
   no name.
   ... My feeling is that since Joanie opened the issue in Github,
   it's her job to close it.
   ... That practice is something I've encountered in other
   projects.
   ... So I will leave it to Joanie to close it, unless she says
   "no"

   RS: I'm fine with the changes.

   JS: What is our practice for closing issues in our group?
   ... Who has the responsibility or the privilege.

   RS: I think the chair of the task force has the right to do it.
   ... Why should we change what works?

   JD: I tend to close issues I am pretty sure I did correctly
   (e.g. with the ARIA spec, as editor).
   ... But I add the comment that if the reporter finds any
   problems with my change, he/she should feel free to reopen.
   ... And it sounds like what you did is correct, though I'm
   scribing and cannot check.
   ... So go ahead and close it.

   JS: Done.

ACTION-1707 (Joseph/All) pending final review.

   <clown> action-1707

   <trackbot> action-1707 -- Joseph Scheuhammer to Edit the AXAPI
   mapping table for roles presentation and none to specify that
   when applied to tables and lists, the children are exposed as
   AXStaticText accessibles. -- due 2016-12-06 -- PENDINGREVIEW

   <trackbot> [15]http://www.w3.org/WAI/ARIA/track/actions/1707

     [15] http://www.w3.org/WAI/ARIA/track/actions/1707

   JS: This was, up until quite recently, just changing a mapping
   in AXAPI.
   ... But it has since turned into something bigger and more
   general.

   <clown>
   [16]https://rawgit.com/w3c/aria/master/core-aam/core-aam.html#r
   ole-map-none

     [16]
https://rawgit.com/w3c/aria/master/core-aam/core-aam.html#role-map-none

   JS: Joanie suggested reversing the text.
   ... Link to the general rule first, and then state the specific
   cases.
   ... That's what I've done (see above URL)
   ... (Reads from spec)

   RS: You're saying empty containers should be pruned?

   JS: Yes.
   ... Anything else that needs to be done here, or can I close
   it?

   JD: Looks good to me. Rich?

   RS: Agreed.

   JS: Closing it then.

(All) Specifying AAPI UIA interface in focus events table

   JS: Rich asked me to do this awhile ago.

   <clown>
   [17]https://rawgit.com/w3c/aria/master/core-aam/core-aam.html#f
   ocus_state_event_table

     [17]
https://rawgit.com/w3c/aria/master/core-aam/core-aam.html#focus_state_event_table

   JS: I had some second thoughts and so wanted to talk more about
   it.
   ... There's a focus state table (URL above).
   ... You (Rich) wanted to know the interface associated with
   what is in the table.
   ... We discussed this further via Skype.
   ... And we decided to follow up with Bogdan to get the
   information.
   ... But then it occurred to me that this issue applies to all
   accessibility APIs; not just UIA.
   ... So do we really want to do this?

   RS: They have a bunch of control patterns in UIA.
   ... Maybe it doesn't mater if Microsoft is going to do manual
   testing.

   JS: But they started the whole manual testing.

   RS: They may have had a change in available resources, but they
   do not have an ATTA.
   ... I'm writing tests and I need to know what tests to write
   for automated testing.
   ... Joanie has provided some mappings for ATK/AT-SPI2.

   JS: But not for the focus table.

   RS: Right, not for the focus table.
   ... UIA has properties and all these different interfaces, so
   I've had to look around to see where things are.

   JS: I can then go ahead and email Bogdan.
   ... But just for the focus table, or for everything?
   ... I'm looking at the mapping for progress bar, and the
   control pattern is progressbar.

   RS: But now you have a range value pattern.

   JS: Yes, it has the range value pattern too, but I don't know
   where that is.

   RS: I'm looking at the test harness for range values.
   ... As an example, scroll bar.

   <richardschwerdtfeger>
   [18]https://www.w3.org/wiki/ARIA_1.1_Testable_Statements#defaul
   t_values:_role.3D.22scrollbar.22

     [18]
https://www.w3.org/wiki/ARIA_1.1_Testable_Statements#default_values:_role.3D.22scrollbar.22

   RS: (Reads from URL above)
   ... I can figure it out for testing purposes.

   JS: I think the person running the test needs to know about the
   patterns and how to access them.

   RS: But it would be nice to have more specific information,
   especially for automation.

   JS: So do you want it for all tables?

   RS: There are already places we do this in the spec.
   ... And I think it's good to be specific.

   JS: I'll ask Bogdan for this information.

(All) update on AX API GH-ISSUEs

   JS: There are a whole bunch of Github issues for AX API
   mappings.
   ... As Joanie goes through and looks for things for WebKitGtk,
   she will tell us if anything is being done for these issues.
   ... And we can update things as we go.

(All) aria-label and the accname-aam

   JS: I've been getting hit with emails and Github issues related
   to the AccName computation.
   ... aria-label was done entirely via email.
   ... I seem to have addressed James Craig's concerns and no one
   has objected.

   <clown> "if the current node has an aria-label attribute whose
   value is not the empty string and does not consist only of
   white space"

   JS: He was concerned about what does "non-empty" really mean?
   ... So I proposed the text I've quoted above.
   ... Anyone object to that change?

   RS: No

   JS: There's a similar statement related to aria-labelledby, and
   a similar concern.

   <clown> Current wording is: "if aria-labelledby is non-empty,
   then …"

   <clown> First suggested change: "if the aria-labelledby
   attribute is non-empty and contains at least on vaild IDREF,
   then …"

   <clown> Second suggested change: "if the aria-labelledby
   attribute contains at least on vaild IDREF, then …"

   JS: Before I make the change, I'll propose it on the Github
   issue and see what he thinks.
   ... Does anyone here have objections to those changes?

   RS: I think it's fine.

(All) aria-labelledby and the accname-aam

   JS: We just talked about this under the previous item.

Triage issues/actions:
[19]https://www.w3.org/WAI/ARIA/track/products/23

     [19] https://www.w3.org/WAI/ARIA/track/products/23

   <clown> actions-1681?

   <clown> action-1681?

   <trackbot> action-1681 -- Joseph Scheuhammer to Propose new
   wording, as an editorial change only to clarify the inclusion
   rules in section 5.1.2 -- due 2016-12-06 -- OPEN

   <trackbot> [20]http://www.w3.org/WAI/ARIA/track/actions/1681

     [20] http://www.w3.org/WAI/ARIA/track/actions/1681

   JS: I'm working on this.
   ... We had a bunch of rules in the spec regarding when role
   presentation is overridden.
   ... Rich added text to the ARIA spec.
   ... I need to make corresponding changes in Core AAM.

   RS: I thought you'd done this already.

   JS: No, but I hope to have it done by next week.

   <clown> action-1732?

   <trackbot> action-1732 -- Michael Cooper to Work on conformance
   section for mappings (note there is a section, but change
   id=conformance so respec picks up) -- due 2015-10-14 -- OPEN

   <trackbot> [21]http://www.w3.org/WAI/ARIA/track/actions/1732

     [21] http://www.w3.org/WAI/ARIA/track/actions/1732

   JS: On action-1732, I'll ping Michael Cooper.

   <clown> action-2056?

   <trackbot> action-2056 -- Joanmarie Diggs to Coordinate the
   mappings for the various AAPIs of the enumerated aria-haspopup
   values -- due 2016-06-07 -- OPEN

   <trackbot> [22]http://www.w3.org/WAI/ARIA/track/actions/2056

     [22] http://www.w3.org/WAI/ARIA/track/actions/2056

   JS: This might be not relevant any more
   ... There's a discussion between James and Rich about
   relationships.
   ... Is this one of the Github issues?
   ... What James is saying is that we don't need this.
   ... (Reads from issue)
   ... But James has not made any other comments in tracker.
   ... I'll check Github.
   ... I don't see a Github issue for this. Why not?

   RS: I don't think we made an issue to this.

   JS: Should I put "unimplemented" in the Core AAM?

   RS: No. Microsoft is doing something similar.

   JS: There are mappings in AXAPI, show-menu and press actions.
   ... Microsoft has the expand/collapse pattern. That's it.
   ... I'm inclined to close this action if it's alright with
   everyone.

   RS: Ok.

   JS: Closed.

   <clown> action-2017?

   <trackbot> action-2017 -- Cynthia Shelly to Look at annotation
   overlap with aria-linktype -- due 2016-03-04 -- CLOSED

   <trackbot> [23]http://www.w3.org/WAI/ARIA/track/actions/2017

     [23] http://www.w3.org/WAI/ARIA/track/actions/2017

   <clown> aciton-2057?

   <clown> action-2057?

   <trackbot> action-2057 -- Joanmarie Diggs to Suggest that axapi
   mapping of role="form" should be a landmark in the AXSubRole.
   -- due 2016-05-03 -- OPEN

   <trackbot> [24]http://www.w3.org/WAI/ARIA/track/actions/2057

     [24] http://www.w3.org/WAI/ARIA/track/actions/2057

   <clown>
   [25]https://rawgit.com/w3c/aria/master/core-aam/core-aam.html#r
   ole-map-form

     [25]
https://rawgit.com/w3c/aria/master/core-aam/core-aam.html#role-map-form

   JD: I've done a number of the WebkitGtk + Safari issues (and in
   some cases implementations).
   ... But right now I cannot recall if I did this or not, and
   would need to review my commits, etc.

   JS: Ok, I'll leave this open and not change the due date.

   JD: Fine.

   <clown> action-2117?

   <trackbot> action-2117 -- Joanmarie Diggs to Update core-aam to
   not expose aria-details relationships that reference elements
   that are hidden. -- due 2016-09-22 -- OPEN

   <trackbot> [26]http://www.w3.org/WAI/ARIA/track/actions/2117

     [26] http://www.w3.org/WAI/ARIA/track/actions/2117

   <clown>
   [27]https://rawgit.com/w3c/aria/master/core-aam/core-aam.html#a
   riaDetails

     [27]
https://rawgit.com/w3c/aria/master/core-aam/core-aam.html#ariaDetails

   RS: The problem is that if you go there and there's nothing
   there to navigate within, you wind up in the aether.

   <clown>
   [28]https://w3c.github.io/aria/aria/aria.html#aria-details

     [28] https://w3c.github.io/aria/aria/aria.html#aria-details

   JS: The spec URL is above.
   ... Looking at the example 18 you have a paragraph that might
   be hidden.
   ... I can see why you wouldn't expose it, because there's
   nothing to expose.
   ... The mappings on all platforms is a pointer of some kind.
   ... So is this as simple as adding a sentence, if the details
   element is not present, do not expose a relationship?

   RS: If the *referenced* details element is not present....

   JS: In the accessibility tree.

   RS: I'm sure the browsers handle this already.
   ... Do we have to write a test for this?

   JS: We probably should.

   <clown> "If the referenced details accessible object is not
   present in the accessibility tree (e..g,. because the details
   are hidden), do not expose {the relationship}"

   JS: Proposed text above.
   ... For each accessibility API, I'd adjust "{the relationship}"
   to reflect that API's mapping.
   ... I'll reassign this issue to me.

   <clown> action-2130

   <trackbot> action-2130 -- Joanmarie Diggs to Determine if AX
   AXPI supports an event if the accessible description changes.
   -- due 2016-12-20 -- OPEN

   <trackbot> [29]http://www.w3.org/WAI/ARIA/track/actions/2130

     [29] http://www.w3.org/WAI/ARIA/track/actions/2130

   JD: I emailed Apple, I need to follow up.
   ... But the short answer is "no".

   JS: I've heard this is handy for tooltips.
   ... Can you (Joanie) update 2130 and then change the owner to
   me?

   JD: Sure.

Summary of Action Items

Summary of Resolutions

   [End of minutes]

Received on Tuesday, 31 January 2017 21:05:21 UTC