[aapi] Minutes: UAI TF Meeting Tue 27 January 2015

Link: http://www.w3.org/2015/01/27-aapi-minutes.html

Plain text follows:

   [1]W3C

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

                               - DRAFT -

           Protocols and Formats Working Group Teleconference
                              27 Jan 2015

   See also: [2]IRC log

      [2] http://www.w3.org/2015/01/27-aapi-irc

Attendees

   Present
          Bryan_Garaventa, Cynthia_Shelly, David_Bolter,
          Jason_Kiss, Joanmarie_Diggs, Joseph_Scheuhammer,
          Rich_Schwerdtfeger

   Regrets
   Chair
          Joseph_Scheuhammer

   Scribe
          joanie

Contents

     * [3]Topics
         1. [4]ACTION-1569: (Joseph/Jason) Update re: status of
            section on AAPI differences.
         2. [5]ACTION-842: (Cynthia) Create a diagram to support
            section 1.2 A11Y vs DOM tree.
         3. [6]ACTION-1104: (Cynthia) How to map aria-describedy
            when CSS display:none.
         4. [7]ACTION-1546/1544: (David/Cynthia) How to expose
            rowgroup role for IA2 and UIA.
         5. [8]ACTION-1373/ISSUE-441: (Cynthia) Test case for
            aria-posinset with only some explicit aria-setsize.
         6. [9]ACTION-1553/ISSUE-692: (Joseph) ATK/AT-SPI mapping
            for role="searchbox".
         7. [10]ACTION-1554/ISSUE-692: (Cynthia) UIA mapping for
            role="searchbox".
         8. [11]ACTION-1553/ISSUE-692: (Joseph) ATK/AT-SPI mapping
            for role="searchbox".
         9. [12]ACTION-1556/ISSUE-692: (David) MSAA/IA2 mapping
            for role="searchbox".
        10. [13]ACTION-1555/ISSUE-692: (James) AX API mapping for
            role="searchbox"
        11. [14]ACTION-1562/ISSUE-693: (Joseph) ATK/AT-SPI mapping
            for role="switch".
        12. [15]ACTION-1564/ISSUE-693: (David) MSAA/IA2 mapping
            for role="switch".
        13. [16]ACTION-1563/ISSUE-693: (James) AX API mapping for
            role="switch".
        14. [17]Reviewing the SVG mapping document
     * [18]Summary of Action Items
     __________________________________________________________

   <trackbot> Date: 27 January 2015

   <scribe> agenda: this

   <joanie> wait

   <joanie> scribe: joanie

ACTION-1569: (Joseph/Jason) Update re: status of section on AAPI
differences.

   <clown> action-1569?

   <trackbot> action-1569 -- Jason Kiss to Create a section that
   describes AAPI differences -- due 2015-03-03 -- OPEN

   <trackbot>
   [19]https://www.w3.org/WAI/PF/Group/track/actions/1569

     [19] https://www.w3.org/WAI/PF/Group/track/actions/1569

   JK: In terms of W3C, I'm on a few working groups.
   ... I am also the representative of the New Zealand government.
   ... I agreed to help work on the Accessibility API mappings
   beginning several years ago.
   ... I've been working on this document, but my time is
   constrained to four hours per week, though I wish it were more.
   ... I don't have in depth knowledge of all the Accessibility
   APIs, but I'm happy to do whatever editorial and other such
   tasks I can.

   <clown> issue-540?

   <trackbot> issue-540 -- Should the UAIG have a section to
   describe the API differences -- open

   <trackbot> [20]https://www.w3.org/WAI/PF/Group/track/issues/540

     [20] https://www.w3.org/WAI/PF/Group/track/issues/540

   <clown>
   [21]http://rawgit.com/w3c/aria/master/accname-aam/accname-aam.h
   tml#accessible-name-and-description-mapping

     [21]
http://rawgit.com/w3c/aria/master/accname-aam/accname-aam.html#accessible-name-and-description-mapping

   JK: The mappings for name and description sort of switched,
   with description being exposed as AX Help.
   ... I raised this issue, and it was explained to me.
   ... But I'm still wondering if the accessible name,
   description, help text should be explained further in the spec.
   ... Another issue is how ATK and AT-SPI related to one another.
   ... I noticed that a note has been added to the Core Mapping
   spec, which is cool.
   ... I'm wondering what the scope of the note I'm working on
   should be.
   ... I'm working on comparing the APIs and looking for any
   potential areas of confusion.
   ... I'm also thinking there may be additional differences of
   interest.
   ... But for me to work on this, I need more input from people
   with knowledge of the specific APIs.
   ... Some of these differences have to do with how user agents
   implement or treat the APIs.
   ... One example is Mozilla's use of a string valie in VARIANT
   in MSAA.

   <jkiss> use of a string value in VARIANT

   <jkiss> in MSAA

   <jkiss>
   [22]http://rawgit.com/w3c/aria/master/html-aam/html-aam.html#us
   e-of-msaa-variant-by-some-user-agents

     [22]
http://rawgit.com/w3c/aria/master/html-aam/html-aam.html#use-of-msaa-variant-by-some-user-agents

   DB: I'm happy that Jason is looking into this, beyond just the
   mappings.

   JK: I just posted a link to the section in the Mapping document
   where this came up.
   ... Cynthia said Mozilla is doing this to expose information
   when it cannot be exposed via other API.

   JS: I read they just use the IA2 role now; not accrole.

   <davidb> Mozilla bug:
   [23]https://bugzilla.mozilla.org/show_bug.cgi?id=798492

     [23] https://bugzilla.mozilla.org/show_bug.cgi?id=798492

   JK: Another issue is the use of the DOM by ATs.

   JS: Certainly true of ChromeVox (use of the DOM).

   JK: That's broader than how the APIs work, however.
   ... Hearing from others about interesting differences worth
   including in a Note would be good.
   ... My interest is that I rely upon these documents and I
   suspect there would be others who look at these documents and
   would find it really helpful to have an explanation of the
   higher-level differences between these APIs.

   JS: Sounds ok to me.

   <clown>
   [24]http://rawgit.com/w3c/aria/master/accname-aam/accname-aam.h
   tml#accessible-name-and-description-mapping

     [24]
http://rawgit.com/w3c/aria/master/accname-aam/accname-aam.html#accessible-name-and-description-mapping

   JS: The URL above is in the accessible name calculation
   document. And the column for AX API is rather lengthy.

   JK: I was looking at that. And for ATK/AT-SPI the document
   suggests that there is not an accessible name property.

   JS: My recollection is that you have to call an atk method

   atk_object_get_name()

   JS: There may be a property there, but you cannot see it.

   <jkiss>
   [25]https://developer.gnome.org/atk/unstable/AtkObject.html#Atk
   Object--accessible-name

     [25]
https://developer.gnome.org/atk/unstable/AtkObject.html#AtkObject--accessible-name

   JK: There (above) is the accessible name property. It's a
   readwrite property.
   ... I couldn't find that in AT-SPI

   <clown>
   [26]https://developer.gnome.org/libatspi/2.13/AtspiAccessible.h
   tml

     [26] https://developer.gnome.org/libatspi/2.13/AtspiAccessible.html

   JD: I will verify that there is not a property for the name.

   JK: Perhaps that's more detailed than is worth it?

   JS: That's more of an AT Vendor question. Do you need that
   level of detail?

   <clown> scribenick: clown

   JD: As an AT developer, I don't go to W3C specs, but I go to
   the AAPI documentation for that platform.

   JK: That makes sense, but it's still worthwhile to me.
   ... I am not a browser implementor, nor an AT developer.
   ... I could just go to all of the AAPI documentation and read
   it.
   ... Or, get a summary of the differences in the W3C documents.

   JD: I was more answering Joseph's question, "what does an AT
   vendor need to know"?
   ... So, keep the audience in mind.

   JK: Yes, sure.
   ... I could draft a section, and include the items I think
   belong.
   ... And then see if others want to add info to that.
   ... And then bring it back to this group.

   JD: I think it's a good idea.
   ... I would be happy to read the draft.

   <joanie> scribe: joanie

   JS: I think the next step is for you to draft it and put it
   Somewhere(tm).

   JK: Where?
   ... I was thinking of adding it under Section 1.1 which
   discusses Accessibility APIs.
   ... Like under a new 1.1.1. That states there are differences.
   ... This would then follow with examples illustrating these
   differences.
   ... I work this up. Should I added it to the latest draft in
   github?

   JS: I'd prefer you branch, either according to the action or
   the issue.
   ... You decide a branch name based on the issue, action, or
   both, to ensure we know what the branch is about.

   JK: I'll do that.

ACTION-842: (Cynthia) Create a diagram to support section 1.2 A11Y vs
DOM tree.

   <clown> action-842?

   <trackbot> action-842 -- Cynthia Shelly to Create a diagram to
   support section 1.2 A11Y vs DOM tree -- due 2014-12-16 -- OPEN

   <trackbot>
   [27]https://www.w3.org/WAI/PF/Group/track/actions/842

     [27] https://www.w3.org/WAI/PF/Group/track/actions/842

   JS: Cynthia, this action came due about two weeks ago.
   ... This is from 2011.

   CS: So Michael Cooper and I made some progrss at TPAC.
   ... I have a hand-drawn version that needs to be turned into
   another format.
   ... I've not yet had an opportunity to do so.
   ... Give me to April 1st or the closest to our meeting.

   JS: March 31st it is.

   CS: I took digital photos too. :)

ACTION-1104: (Cynthia) How to map aria-describedy when CSS
display:none.

   <clown> action-1104?

   <trackbot> action-1104 -- Cynthia Shelly to Define what the
   accessibility API mapping is for UIA on aria-describedby in
   section 5.5.1 table when the element does not exist in the
   accessibility tree such as when css: display:none applies --
   due 2015-01-14 -- OPEN

   <trackbot>
   [28]https://www.w3.org/WAI/PF/Group/track/actions/1104

     [28] https://www.w3.org/WAI/PF/Group/track/actions/1104

   JS: You got some emails about this (Cynthia).

   <clown>
   [29]http://rawgit.com/w3c/aria/master/accname-aam/accname-aam.h
   tml#accessible-name-and-description-mapping

     [29]
http://rawgit.com/w3c/aria/master/accname-aam/accname-aam.html#accessible-name-and-description-mapping

   JS: That mapping table I brought up before (URL above), if you
   look at UIA, there's nothing explaining what you do if the
   objects are not in the accessibility tree.

   CS: This is an ongoing discussion. I need to follow up on this.
   Please give me two weeks.

   JS: February 10th.

   JK: You noted that UIA doesn't explain what to do if it's not
   in the accessibility tree. It seems to me that this is lacking
   for others as well.

   JS: Not for IA2/ATK.
   ... So IA2 and ATK have both a way to provide the string and to
   provide the object.
   ... UIA doesn't have a means to provide the string.

   CS: I'm trying to convince people we need this means.

ACTION-1546/1544: (David/Cynthia) How to expose rowgroup role for IA2
and UIA.

   <clown> action-1546?

   <trackbot> action-1546 -- David Bolter to Make sure the
   rowgroup mapping for msaa/ia2 is accurate -- due 2014-12-23 --
   OPEN

   <trackbot>
   [30]https://www.w3.org/WAI/PF/Group/track/actions/1546

     [30] https://www.w3.org/WAI/PF/Group/track/actions/1546

   <clown> action-1544?

   <trackbot> action-1544 -- Cynthia Shelly to Document rowgroup
   mappings for ie 11 for both uia express and uia -- due
   2014-12-23 -- OPEN

   <trackbot>
   [31]https://www.w3.org/WAI/PF/Group/track/actions/1544

     [31] https://www.w3.org/WAI/PF/Group/track/actions/1544

   JS: GNOME decided to not expose that role at all.

   CS: I think we do something with it.

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

     [32]
http://rawgit.com/w3c/aria/master/core-aam/core-aam.html#role-map-rowgroup

   JS: The current mappings are above.
   ... Apple doesn't map it either.
   ... And IA2 exposes it as role group, and UIA as group.

   RS: Firefox put it in there, but we're not sure why.
   ... I don't think it's a show stopper, because I don't think
   the ATs pay much attention to it.

   CS: I haven't had a chance to look. Please give me a week.

ACTION-1373/ISSUE-441: (Cynthia) Test case for aria-posinset with
only some explicit aria-setsize.

   <clown> action-1373?

   <trackbot> action-1373 -- Cynthia Shelly to Create a test case
   for when aria-posinset and aria-setsize are provided explicitly
   on some but not all elements. -- due 2014-12-16 -- OPEN

   <trackbot>
   [33]https://www.w3.org/WAI/PF/Group/track/actions/1373

     [33] https://www.w3.org/WAI/PF/Group/track/actions/1373

   CS: This one's harder. Give me four weeks.

   JS: It's writing a test case.

   CS: It's more than one test case; it's a bunch of different
   scenarios.
   ... I know there are bugs, which is why I made that action in
   the first place.
   ... Four weeks is CSUN, so why don't we make it three.

ACTION-1553/ISSUE-692: (Joseph) ATK/AT-SPI mapping for
role="searchbox".

   JS: This is more of an update.

   <clown> action-1553?

   <trackbot> action-1553 -- Joseph Scheuhammer to Provide
   ATK/AT-SPI2 mapping for role="searchbox" -- due 2015-01-21 --
   CLOSED

   <trackbot>
   [34]https://www.w3.org/WAI/PF/Group/track/actions/1553

     [34] https://www.w3.org/WAI/PF/Group/track/actions/1553

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

     [35]
http://rawgit.com/w3c/aria/master/core-aam/core-aam.html#role-map-searchbox

   JS: We discussed this last week. We have a new role in ARIA:
   searchbox.
   ... And we got the mappings for ATK/AT-SPI. And I managed to
   change the document to add the role, but only ATK/AT-SPI has a
   mapping. I still need them for the other platform mappings for
   this role.

   <clown> action-1554?

   <trackbot> action-1554 -- Cynthia Shelly to Provide UIA mapping
   for role="searchbox" -- due 2015-01-21 -- OPEN

   <trackbot>
   [36]https://www.w3.org/WAI/PF/Group/track/actions/1554

     [36] https://www.w3.org/WAI/PF/Group/track/actions/1554

ACTION-1554/ISSUE-692: (Cynthia) UIA mapping for role="searchbox".

   CS: Give me two weeks so that I can verify how to map it.
   ... Is this a landmark?

   JS: No, search is a landmark.

   RS: This is an entry.

   JS: This is why ATK mapped it this way.
   ... But as I understand it, they are going to expand their
   roles.

   RS: The easy way to do this is map it to a text field and then
   have ATs look at an attribute to see if it's a searchbox.

   (Group discusses object attributes)

ACTION-1553/ISSUE-692: (Joseph) ATK/AT-SPI mapping for
role="searchbox".

   <clown> action-1553?

   <trackbot> action-1553 -- Joseph Scheuhammer to Provide
   ATK/AT-SPI2 mapping for role="searchbox" -- due 2015-01-21 --
   CLOSED

   <trackbot>
   [37]https://www.w3.org/WAI/PF/Group/track/actions/1553

     [37] https://www.w3.org/WAI/PF/Group/track/actions/1553

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

     [38]
http://rawgit.com/w3c/aria/master/core-aam/core-aam.html#role-map-switch

ACTION-1556/ISSUE-692: (David) MSAA/IA2 mapping for role="searchbox".

ACTION-1555/ISSUE-692: (James) AX API mapping for role="searchbox"

ACTION-1562/ISSUE-693: (Joseph) ATK/AT-SPI mapping for role="switch".

   JS: ATK/AT-SPI2 mapped it to toggle button.

ACTION-1564/ISSUE-693: (David) MSAA/IA2 mapping for role="switch".

   RS: I'd do this as an object attribute just like with
   searchbox.

   <clown> action-1563?

   <trackbot> action-1563 -- James Craig to Provide AX API mapping
   for role="switch" -- due 2015-01-22 -- OPEN

   <trackbot>
   [39]https://www.w3.org/WAI/PF/Group/track/actions/1563

     [39] https://www.w3.org/WAI/PF/Group/track/actions/1563

   CS: I'll take a look at this.

ACTION-1563/ISSUE-693: (James) AX API mapping for role="switch".

   <richardschwerdtfeger>
   [40]http://rawgit.com/w3c/aria/master/svg-aam/svg-aam.html

     [40] http://rawgit.com/w3c/aria/master/svg-aam/svg-aam.html

Reviewing the SVG mapping document

   RS: The SVG accessibility taskforce has gone through this
   already.
   ... I think it would be good if this group could look at it as
   well.

   JS: What sort of time frame?

   RS: It's a first working draft. I don't expect the comments to
   be thorough.
   ... If you have anything significant, I'd love to hear it.
   ... We want to get this one and the HTML one out in February.

   JK: I think it's possible. I'm unavailable for a couple of
   weeks in February, but I think the HTML document will be ready.
   ... Some addition fixes to leverage off of the core, similar to
   what SVG is doing, might be useful.

   JS: I still object to the use of none.
   ... SVG animate element, for example.
   ... Do you mean the actual "none" role, or that there is no
   role?

   RS: There is a "none" role and a none (for no role).
   ... animate doesn't get mapped at all.

   JK: It might be mistakenly linked to role="none"

   <cyns> have to run to another meeting

   JK: Should it say "no role"?

   RS: It could, but it is essentially the same.

   JS: To me that means it's there for presentation purposes.

   JK: To avoid the ambiguity....

   RS: But it has the same effect.

   (Discussion of validity of mapping it to a pane)

   RS: I will raise an issue.

   JS: There is a possibility that something with role="none" can
   gain focus.

   RS: But that's an error.

   JS: It's an author error to put the presentational role on
   something which is focusable.

   RS: Every single drawing is persistent in memory. Can you
   imagine if you had an accessible object for every single item?
   The tree would be huge.

   JS: Could an author put a tab index on an SVG element so that
   an AT could walk through the parts of the diagram?

   RS: Yes, but you'd want to put it on an element.
   ... It shouldn't be mapped, but if it becomes focusable then
   you have to?

   JS: Yes.

   RS: Then I have to put something in the spec because it's an
   error condition.
   ... I still want this to be none, which means there should be
   no mapping.
   ... I think what we have to do is have rules about mapping
   elements with tab index.
   ... I will make an editorial note.

   (Group discusses the specifics)

   <clown> <img role="presentaion">

   <clown> <img role="presntation" tabindex="0">

   JS: the first of the above won't be mapped; the second will.
   ... My read is that IF it needs to be in the accessibility
   tree, treat it as a pane. Otherwise don't worry about it.

   RS: I'm going to make a note and create an issue for my spec.

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

     [41]
http://rawgit.com/w3c/aria/master/core-aam/core-aam.html#include_elements

   JS: The exceptions to not exposing it are described at the URL
   above.

Summary of Action Items

   [End of minutes]

Received on Tuesday, 27 January 2015 21:20:27 UTC