[aapi] Minutes: UAI TF Meeting Tue Feb 23, 2016

Link: https://www.w3.org/2016/02/23-aapi-minutes.html

Plain text follows.

   [1]W3C

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

   Accessible Rich Internet Applications Working Group Teleconference

23 Feb 2016

   See also: [2]IRC log

      [2] http://www.w3.org/2016/02/23-aapi-irc

Attendees

   Present
          Joanmarie_Diggs, Bryan_Garaventa, Joseph_Scheuhammer,
          Cynthia_Shelly, Rich_Schwerdtfeger

   Regrets
   Chair
          Joseph_Scheuhammer

   Scribe
          joanie

Contents

     * [3]Topics
         1. [4]ACTION-1569/1585/ISSUE-540 (Joseph) Add to section
            on AAPI differences highlighting the features of UIA
         2. [5]ACTION-1593 (Cynthia) Compare implementations of
            AAPI managed states and look for gaps.
         3. [6]ACTION-1531 (All) IA2 mapping of role="text" --
            using ROLE_SYSTEM_STATICTEXT in editor's draft.
         4. [7]ACTION-1691/ISSUE-1010 (Joseph) UIA mapping for
            aria-errormessage and role="error".
         5. [8]Providing feedback and changes on Core AAM mappings
         6. [9]ACTION-1691/ISSUE-1010 (Joseph) UIA mapping for
            aria-errormessage and role="error".
         7. [10]Comboboxes
         8. [11]ACTION-1541 (Joseph) AXAPI mapping for aria-modal
            property.
     * [12]Summary of Action Items
     * [13]Summary of Resolutions
     __________________________________________________________

   <clown> agenda: this

   <scribe> scribe: joanie

ACTION-1569/1585/ISSUE-540 (Joseph) Add to section on AAPI
differences highlighting the features of UIA

   <clown> action-1569

   <trackbot> action-1569 -- Joseph Scheuhammer to Create a
   section that describes AAPI differences -- due 2016-02-23 --
   PENDINGREVIEW

   <trackbot> [14]http://www.w3.org/WAI/ARIA/track/actions/1569

     [14] http://www.w3.org/WAI/ARIA/track/actions/1569

   JS: There's been some work going back almost a year.
   ... Jason Kiss and Cynthia added stuff to the Core AAM
   describing differences between the AAPIs.

   <clown>
   [15]http://rawgit.com/w3c/aria/issue-540/core-aam/core-aam.html
   #comparing-accessibility-apis

     [15]
http://rawgit.com/w3c/aria/issue-540/core-aam/core-aam.html#comparing-accessibility-apis

   JS: I've incorporated their work. See above URI.
   ... What he did... There was previously a note about
   ATK/AT-SPI. Jason took this out of the note.
   ... Jason also added a section describing how different AAPIs
   handle names and descriptions.

   <clown>
   [16]http://rawgit.com/w3c/aria/master/core-aam/core-aam.html#co
   mparing-accessibility-apis

     [16]
http://rawgit.com/w3c/aria/master/core-aam/core-aam.html#comparing-accessibility-apis

   JS: Correction, the URI immediately above is the correct one.
   ... Cynthia's text follows next.

   <clown>
   [17]http://rawgit.com/w3c/aria/master/core-aam/core-aam.html#in
   tro_aapi

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

   JS: Some of Cynthia's text I put up higher in the section.
   ... If you go to the intro (URI above), there is a numbered
   list which is also Cynthia's text.
   ... It provides a high-level explanation of what goes into an
   Accessibility API.

   CS: This looks ok to me, though my button example (in 1.3.2)
   could use some work.
   ... Please delete the sentence with the example. I'll provide
   another.

   <scribe> ACTION: Cynthia to provide a new example for the
   Accessibility API explanatory text. [recorded in
   [18]http://www.w3.org/2016/02/23-aapi-minutes.html#action01]

     [18] http://www.w3.org/2016/02/23-aapi-minutes.html#action01]

   <trackbot> Created ACTION-2032 - Provide a new example for the
   accessibility api explanatory text. [on Cynthia Shelly - due
   2016-03-01].

   <clown> action-2032

   <trackbot> action-2032 -- Cynthia Shelly to Provide a new
   example for the accessibility api explanatory text. -- due
   2016-03-01 -- OPEN

   <trackbot> [19]http://www.w3.org/WAI/ARIA/track/actions/2032

     [19] http://www.w3.org/WAI/ARIA/track/actions/2032

   JS: I'll wait for Jason to get back to me. I'll ping him again.
   ... Other comments?

   CS: I'm happy to be making progress on this one. :)

   Group agrees

ACTION-1593 (Cynthia) Compare implementations of AAPI managed states
and look for gaps.

   <clown> action-1593

   <trackbot> action-1593 -- Cynthia Shelly to Compare
   implementations of AAPI managed states and look for gaps. --
   due 2016-02-23 -- OPEN

   <trackbot> [20]http://www.w3.org/WAI/ARIA/track/actions/1593

     [20] http://www.w3.org/WAI/ARIA/track/actions/1593

   JS: This is another AAPI related one.

   CS: I've not yet had time to do this one. Leave it overdue.

ACTION-1531 (All) IA2 mapping of role="text" -- using
ROLE_SYSTEM_STATICTEXT in editor's draft.

   <clown> action-1531?

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

   <trackbot> [21]http://www.w3.org/WAI/ARIA/track/actions/1531

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

   JS: This should be a quick one.
   ... I finally got around to changing the IA2 mapping for
   role="text".

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

     [22]
http://rawgit.com/w3c/aria/master/core-aam/core-aam.html#role-map-text

   JS: URI is above.
   ... You'll see it says ROLE_SYSTEM_STATICTEXT now.
   ... Any other changes needed?

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

     [23]
http://rawgit.com/w3c/aria/master/core-aam/core-aam.html#role-map-text

   JD: My platform is fine.

   CS: It's fine for mine, although my concerns about this role
   remain.

   RS: I did get Steve to limit the usage of role="text".
   ... But James N. has concerns about that.

ACTION-1691/ISSUE-1010 (Joseph) UIA mapping for aria-errormessage and
role="error".

Providing feedback and changes on Core AAM mappings

   CS: I have a number of changes to propose. How would you like
   them?

   JS: I would prefer single issues over a single large one. And
   tracker over github.

   CS: I can do issues, but I'm trying to reduce overhead.

   JS: Ok, you could do one big pull request.

   CS: Since they are about UIA, I don't anticipate concerns or
   discussions regarding my proposed changes.

   <clown> action-1691

   <trackbot> action-1691 -- Joseph Scheuhammer to Propose UIA
   mapping for aria-errormessage. -- due 2015-12-31 --
   PENDINGREVIEW

   <trackbot> [24]http://www.w3.org/WAI/ARIA/track/actions/1691

     [24] http://www.w3.org/WAI/ARIA/track/actions/1691

ACTION-1691/ISSUE-1010 (Joseph) UIA mapping for aria-errormessage and
role="error".

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

     [25]
http://rawgit.com/w3c/aria/master/core-aam/core-aam.html#ariaErrorMessage

   JS: All I changed here is (he reads previous and new text)
   ... This is for UIA.

   RS: Do we need a role="error" any more?

   JS: Good question. When I saw the action, it had role="error"
   in it.

   RS: You could handle this in the underlying platform.

   CS: It seems analogous to label.

   RS: As long as it only takes a single ID, in the underlying
   platform you could have a platform role.

   CS: One reason I like it in the markup is for HTML readability.

   RS: I do think we have to address this on the other platforms.

   CS: I think I would point this to an element with a
   role="alertdialog".

   JD: I don't think it's necessarily a dialog.

   CS: Agreed, but this is one pattern that seems sane and a good
   WCAG technique.
   ... Because an error dialog is typical.
   ... I think connecting this to live regions also makes sense.

   JS: Doesn't the appearance of an alertdialog trigger a
   notification?

   JD: On my platform it's supposed to.

   JS: Do I have anything in the Core AAM regarding events for
   aria-errormessage?
   ... I don't. Although it's supposed to be used with
   aria-invalid.
   ... And aria-invalid incurs a state change.
   ... So I don't think we need an issue.

   RS: Yes, it incurs a state change.
   ... But aria-invalid will be on the source.

   JS: We already have in the spec that if the error message is
   pertinent, then aria-invalid should be set to true.
   ... And that should result in a notification.

   RS: Unless it's always been there.

   JS: We have this already in the mapping spec to fire
   notifications in ATK and IA2, but not for UIA.

   RS: But they must have one.

   JS: When a node is created... No. When it's removed, yes.

   RS: I think they should probably fix that.

   JS: Do we need an AAPI for that for aria-errormessage?

   <clown> Issue: is there a need for an AAPI event for
   aria-errormessage

   <trackbot> Created ISSUE-1013 - Is there a need for an aapi
   event for aria-errormessage. Please complete additional details
   at <[26]http://www.w3.org/WAI/ARIA/track/issues/1013/edit>.

     [26] http://www.w3.org/WAI/ARIA/track/issues/1013/edit

   RS: Creating events are more expensive....

   JS: I'll raise the issue and we can discuss it.

Comboboxes

   BG: I don't understand why we're taking something that's
   accessible and creating something that's new that has no
   mappings and is not accessible.

   RS: The original issue was the need to expand comboboxes to
   more popup types than listboxes.
   ... Matt wanted to add things like grid and tree and dialog.
   ... If you have a popup.... Freedom Scientific said it better
   be modal.
   ... Because when you bring up a dialog box, you can be Tabbing
   through a dialog box and then be in the middle of nowhere.

   BG: I agree. I'm opposed to adding dialog.

   RS: Removing dialog would make things easier.
   ... But you still have other popups.
   ... Are they going to be modal or not?
   ... Example, grids are rather large.
   ... What do you want/expect when you press Tab at the end of a
   grid?
   ... Do we need to specify a modal property?
   ... For a tree widget, it's pretty straightforward I think.
   ... But a grid....
   ... Matt is a power user. But not all users are. And what
   happens if you exit the grid without expecting that?
   ... Then you don't know where you are.

   BG: I understand the problem.
   ... But ... Basically when you're dealing with screen readers,
   the interaction is pretty granular.
   ... You're dealing with the object that has focus.
   ... But in a combo box, you're only interacting with the
   combobox or not.
   ... The problem with dialog is that once you're in a dialog you
   are no longer in the combobox.
   ... That's fine if they want to create it like that.
   ... You don't need to put that in the spec, however.
   ... I have the same problem with grid and tree.
   ... Years ago, people started making comboboxes editable
   because it made sense and was easy.
   ... But if we create a brand new structure that has to be
   remapped, it's going to cause interoperability issues.
   ... And make my life quite difficult.

   RS: I understand what Matt is trying to do, but...
   ... The difference between 1.0 and 1.1 is the relationship
   changed from controls to owns.
   ... What I would like to be able to do is say: If you are going
   to make such a UI object, it needs to have a popup.
   ... aria-haspopup would make it possible to indicate this is a
   different type of interaction.
   ... But when you get to the target, you have to know what the
   popup means.
   ... Is it really closing the combobox or not?
   ... I think it would be better to make it modal so that people
   don't get lost.
   ... I'm worried about having a much more complex window being
   brought up.
   ... And people getting very confused.
   ... This is why I've been pushing modal.

   BG: Our clients have many different combobox implementations.
   ... And usually the edit field is the only component that has
   focus.
   ... And aria-activedescendant to point to the selected item,
   e.g. in the grid.
   ... And this works.
   ... So the changes proposed strike me as more confusing.
   ... And also strike me as an APG thing rather than a spec
   thing.

   RS: There are a lot of people with actions outstanding.
   ... And we need to get ARIA 1.1 done.
   ... And I don't want to consume all of our time with one role.

ACTION-1541 (Joseph) AXAPI mapping for aria-modal property.

   <clown> aciton-1541

   <clown> action-1541

   <trackbot> action-1541 -- Joseph Scheuhammer to Investigate and
   propose mappings for aria-modal property for axapi. -- due
   2016-02-08 -- OPEN

   <trackbot> [27]http://www.w3.org/WAI/ARIA/track/actions/1541

     [27] http://www.w3.org/WAI/ARIA/track/actions/1541

   JS: I thought I had this nailed down.
   ... This is putting into the mapping spec AX API's mapping for
   aria-modal.

   <clown> [28]https://bugs.webkit.org/show_bug.cgi?id=138566

     [28] https://bugs.webkit.org/show_bug.cgi?id=138566

   JS: There's a fixed bug in WebKit (about URI).
   ... But I cannot find a property that they set.
   ... I was wondering if Joanie could have a look.

   JD: I'll look.
   ... My guess is that there is already an isModal() being called
   in the existing Mac platform code within WebKit.
   ... And because of that, there was not a need to include
   anything looking like a new mapping in the patch associated
   with that bug.
   ... But I can look.

   (Group looks using Accessibility Inspector in OS X)

   JS: I'll ping James Craig and ask him what I should put in the
   table for AX API for aria-modal.

Summary of Action Items

   [NEW] ACTION: Cynthia to provide a new example for the
   Accessibility API explanatory text. [recorded in
   [29]http://www.w3.org/2016/02/23-aapi-minutes.html#action01]

     [29] http://www.w3.org/2016/02/23-aapi-minutes.html#action01

Summary of Resolutions

   [End of minutes]

Received on Tuesday, 23 February 2016 21:08:11 UTC