[aapi] Minutes: UAI TF Meeting Tue 18 November 2014 (Take 2)

(Turns out I had done 'scribnick' instead of 'scribenick' mid meeting.
This is now corrected, including in the plain text version below.
Apologies for the noise. --jd)

Link: http://www.w3.org/2014/11/18-aapi-minutes.html

Plain text follows:
   [1]W3C

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

                               - DRAFT -

           Protocols and Formats Working Group Teleconference
                              18 Nov 2014

   See also: [2]IRC log

      [2] http://www.w3.org/2014/11/18-aapi-irc

Attendees

   Present
          Joseph_Scheuhammer, +1.416.848.aaaa, David_Bolter,
          Joanmarie_Diggs, Bryan_Garaventa, Cynthia_Shelly

   Regrets
          Rich_Schwerdtfeger

   Chair
          Joseph_Scheuhammer

   Scribe
          joanie, clown

Contents

     * [3]Topics
         1. [4]ISSUE-587/ACTION-1528: (David) Investigate the
            proper ia2 mappings for aria-current.
         2. [5]Do we meet next week?
         3. [6]ISSUE-587/ACTION-1527: (Cynthia) Investigate the
            proper uia mappings for aria-current.
         4. [7]ISSUE-587/ACTION-1529: (Joseph) Investigate the
            proper atk/at-spi mappings for aria-current.
         5. [8]ISSUE-587: (All) discuss use of events when
            aria-current changes.
         6. [9]ISSUE-587/ACTION-1527: (Cynthia) Investigate the
            proper uia mappings for aria-current.
     * [10]Summary of Action Items
     __________________________________________________________

   <trackbot> Date: 18 November 2014

   <clown> agenda: this

   <dalek> exterminate

   <joanie> scribenick: joanie

ISSUE-587/ACTION-1528: (David) Investigate the proper ia2 mappings
for aria-current.

   <clown> action-1528?

   <trackbot> action-1528 -- David Bolter to Bolter to investigate
   the proper ia2 mappings for aria-current -- due 2014-11-18 --
   OPEN

   <trackbot>
   [11]https://www.w3.org/WAI/PF/Group/track/actions/1528

     [11] https://www.w3.org/WAI/PF/Group/track/actions/1528

   JS: There's a new aria-current in the spec.
   ... Briefly, if you're in a navigational structure like a TOC
   off to the side, your current location is your place in the
   TOC.
   ... You could also use it in a wizard or other step-by-step
   process.

   DB: Do we have a status role for that?

   JS: We're not changing the role.

   DB: I'm wondering if this could be done in another way.

   JS: The GNOME people are going to go with ATK_STATIC_ACTIVE

   [12]https://bug740044.bugzilla-attachments.gnome.org/attachment
   .cgi?id=290571

     [12]
https://bug740044.bugzilla-attachments.gnome.org/attachment.cgi?id=290571

   <clown> scribenick: clown

   JD: reads the docs on ATK_STATE_ACTIVE

   Indicates a window is currently the active window, or an object
   is the active subelement within a container or table.
   ATK_STATE_ACTIVE should not be used for objects which have
   ATK_STATE_FOCUSABLE or ATK_STATE_SELECTABLE: Those objects
   should use ATK_STATE_FOCUSED and ATK_STATE_SELECTED
   respectively. ATK_STATE_ACTIVE is a means to indicate that an
   object which is not focusable and not selectable is the
   currently-active item within its parent container.

   DB: regarding this change, you could still have a clickable
   thing be an ACTIVE thing?

   JD: Sure.

   DB: When you say, "objects should use ATK_STATE_FOCUSED and
   ATK_STATE_SELECTED respectively". Is that and/or or both?

   JD: It can be either. Something can be unselected in a
   mult-select list box, but still focussed.

   DB: does ATK have a multi-selected state?

   JD: multi-selecatable.
   ... The new definition is to anticipate a complaint that isn't
   aria-current just focused?

   DB: who want aria-current, and what's it for?

   JD: Here's my use case.
   ... <describes ordering online>
   ... <describles a wizard step-by-step process>

   BG: Let me clarify. The primary reason for aria-current is
   there are situations on the web where an element is marked as
   current in a collection.
   ... It could be a list of items, or a group of images, or a
   list of dates.
   ... Also pagination links, where one is current.
   ... There is not textual way to indicate which one is current.
   ... Sighted users can see some sort of highlighting.

   <joanie> scribenick: joanie

   JS: I'm looking for the aria-current spec text.

   DB: I'm going through the process of, whenever you see a
   proposed aria attribute:

   <clown>
   [13]http://lists.w3.org/Archives/Public/public-pfwg/2014Nov/003
   9.html

     [13] http://lists.w3.org/Archives/Public/public-pfwg/2014Nov/0039.html

   DB: 1) Is it really needed?
   ... 2) Is it as powerful as it can get?
   ... I think we have met 1.
   ... [Describes example in which a live region wouldn't work]

   JS: aria-current has nothing to do with focus
   ... Go back to Joanie's shopping cart example.
   ... You're not focused on the step.

   DB: I totally get that.
   ... The use case I'm trying to imagine is a web page in which
   you might have one or more aria-currents on a page.
   ... I'm thinking of a dynamic page in which one or more
   aria-current's change.
   ... So my example -- bad example -- is the colors of a legend
   changing depening on what country you selected.

   JS: The one (example) that was originally proposed was for
   aria-selected.
   ... But that was rejected as being different.

   DB: A screen reader has to know if aria-current changes for
   some reason.

   BG: I don't think that's needed necessarily.
   ... The author can surround nested regions with labels if
   needed.
   ... It's obvious from the context that the element refers to
   that region.
   ... So I don't think it's necessary to expose anything other
   than that the element is current.

   DB: Is there a situation in which some screen reader notices
   this is aria-current. Does it need to know the relationship
   between it and what makes it current?

   BG: I don't think so. Dominic from Google mentioned it in the
   context in of shadow DOM.

   JS: I mentioned aria-currentfor if you have multiple
   aria-currents on the page
   ... But Matt, James, and others said it wasn't needed.

   DB: Is it expected that you'll only have one?

   JS: No, as far as I know that is not necessarily expected.

   BG: [Agrees] but I don't think that's a problem.

   JS: [Describes example of page with multiple portlets]

   BG: I think it depends on context.

   JS: I remember others saying that too.

   DB: I'm just wondering about getting people to use it.
   ... It has to be clear how to use it.

   JS: True.
   ... aria-currentfor got dropped; aria-current is all we need.

   BG: Cynthia said she was going to try to put it in UIA in
   January, I think

   JS: Is there something similar to STATE_ACTIVE in IA2?
   ... ATK and IA2 are pretty similar.

   DB: Does ATK have STATE_ARMED? Or was that Java?

   JS: Java

   JD: Actually ATK too, but it's stupid and I want it deprecated.

   JS: The "it" there is ambiguous.

   JD: [Makes no comment]

   DB: So there's a STATE_ACTIVE, but it's about the active
   window.

   <clown>
   [14]http://accessibility.linuxfoundation.org/a11yspecs/ia2/docs
   /html/_accessible_states_8idl.html#aafecbfb195aaf7050dc8f75a783
   3bd5fabf466265c1f5b45dd53c86deb28220bc

     [14]
http://accessibility.linuxfoundation.org/a11yspecs/ia2/docs/html/_accessible_states_8idl.html#aafecbfb195aaf7050dc8f75a7833bd5fabf466265c1f5b45dd53c86deb28220bc

   <clown> scribenick: clown

   JD: Years ago, ACTIVE was just for windows.
   ... But then, it was modified to include "sub-elements">
   ... We fixed the docs for ATK to clarify it.
   ... Couldn't we do the same for IA2 docs?
   ... It would mean little change to the IA2 implementation.

   Note: IA2 documentations for STATE_ACTIVE is: "Indicates a
   window is currently the active window, or is an active
   subelement within a container or table."

   <joanie> scribenick: joanie

   JS: We've been discussing this for 1/2 an hour now.
   ... David should ask Alex Surkov about using STATE_ACTIVE in
   IA2 for this.

   DB: Agreed.
   ... I think I'll assign the action to him.

   JS: The current due date is today. Next tuesday?

   DB: Sure.

Do we meet next week?

   JS: In the States it's Thanksgiving Holiday week.
   ... I'll be here. David will be here.

   JD: I will be here.

   BG: I'll be around on Tuesday.

   RESOLUTION: We'll have a meeting next week as usual.

   DB: When do we break for holidays?

   JS: The 16th might be iffy for me, though I'm not sure. I'll
   definitely be out on the 23rd and 30th.
   ... The above refer to December.

ISSUE-587/ACTION-1527: (Cynthia) Investigate the proper uia mappings
for aria-current.

   JS: Cynthia is not here

   <clown> action-1529?

   <trackbot> action-1529 -- Joseph Scheuhammer to Investigate the
   proper atk/at-spi mappings for aria-current -- due 2014-11-18
   -- OPEN

   <trackbot>
   [15]https://www.w3.org/WAI/PF/Group/track/actions/1529

     [15] https://www.w3.org/WAI/PF/Group/track/actions/1529

   JS: This is mine because it was originally Joanie's and she's
   done the work.
   ... Given her notes, I'm ready to add the mappings.
   ... I guess I need to wait until the other platforms.

   <davidb> JD: what about Mac?

   <clown> ACTION: James Craig to investigate the AXAPI mappings
   for aria-current. [recorded in
   [16]http://www.w3.org/2014/11/18-aapi-minutes.html#action01]

   <trackbot> 'James' is an ambiguous username. Please try a
   different identifier, such as family name or username (e.g.,
   jcraig, jhawkins2, jnurthen).

   <clown> ACTION: jcraig to investigate the AXAPI mappings for
   aria-current. [recorded in
   [17]http://www.w3.org/2014/11/18-aapi-minutes.html#action02]

   <trackbot> Created ACTION-1535 - Investigate the axapi mappings
   for aria-current. [on James Craig - due 2014-11-25].

   JS: [Creates new action for AX API mappings]

   <clown> action-1535?

   <trackbot> action-1535 -- James Craig to Investigate the axapi
   mappings for aria-current. -- due 2014-11-25 -- OPEN

   <trackbot>
   [18]https://www.w3.org/WAI/PF/Group/track/actions/1535

     [18] https://www.w3.org/WAI/PF/Group/track/actions/1535

ISSUE-587/ACTION-1529: (Joseph) Investigate the proper atk/at-spi
mappings for aria-current.

ISSUE-587: (All) discuss use of events when aria-current changes.

   JS: When Joanie passed back action-1529, I wondere if
   aria-current should spawn an event.
   ... In which case we need to add this to the event mapping
   table.
   ... I brought this up yesterday at the meeting.
   ... James Craig thought we should make it a "MAY"

   <clown> scribenick: clown

   JS: ATK is going to just make such an event.

   JD: That's not quite right.

   <joanie> action-1529

   <trackbot> action-1529 -- Joseph Scheuhammer to Investigate the
   proper atk/at-spi mappings for aria-current -- due 2014-11-18
   -- OPEN

   <trackbot>
   [19]https://www.w3.org/WAI/PF/Group/track/actions/1529

     [19] https://www.w3.org/WAI/PF/Group/track/actions/1529

   JD: Well, there's the events table (i.e. ARIA), and there's
   what is expected when the state of an accessible object changes
   in <insert any ATK implementation here, including desktop
   applications and toolkits>.
   ... With respect to the latter, the expectation has always been
   that as a general rule, when the state changes, a state-change
   notification is expected.
   ... Otherwise, how is an AT on the platform to know?
   ... It says in the ATK documentation that implementors are
   expected to provide notiifcations of state changes.
   ... ATK doesn't magically do the event. It is still expected
   that the implementor will do the event.

   <joanie> scribenick: joanie

   JS: When I said "ATK is going to do this anyway" I really meant
   "A proper ATK implementation is going to do this anyway."
   ... James Craig wants it to be a MAY because we currently have
   no consensus that it's needed or wanted by any AT.
   ... Cynthia wanted it to be a SHOULD because she wanted it to
   be stronger.
   ... I will put it in as a MAY for AX API and a MUST for ATK.
   But that means we have different normative requirements for
   platforms.
   ... I'm not sure what to do for UIA.
   ... Cynthia wants a SHOULD.
   ... James Craig went on to say that if this is useful and we
   all agree, we can then change it to a MUST.
   ... In other words he's being cautious.

ISSUE-587/ACTION-1527: (Cynthia) Investigate the proper uia mappings
for aria-current.

   JS: We've been talking about mappings for aria-current.
   ... We pretty much have ATK finished.

   <clown> action-1527

   <trackbot> action-1527 -- Cynthia Shelly to Investigate the
   proper uia mappings for aria-current -- due 2015-01-31 -- OPEN

   <trackbot>
   [20]https://www.w3.org/WAI/PF/Group/track/actions/1527

     [20] https://www.w3.org/WAI/PF/Group/track/actions/1527

   JS: David is going to follow up with Surkov on IA2.
   ... What should we use for UIA?

   CS: Selection -- it would become a selected item.
   ... Which means that it needs to have -- there's a container
   that's required.
   ... I put notes in the bug.
   ... The closest thing we have is selected.
   ... But that means we need to have a container.
   ... I think this is related to what Joanie was talking about
   the other day on the call.
   ... About traversing the tree, etc.
   ... I could ask for a new UIA API, but I'd need additional
   justification for why this isn't selection.

   BG: It is possible to have something with a current and a
   selected at the same time. And they wouldn't be the same thing.

   CS: What's the difference?

   BG: If you have a tree, and the current item in the tree is a
   book and there is a current page.

   CS: That sounds like the difference between focus and selection

   BG: When you're arrowing through it, you can have something
   selected and not be current at the same time.
   ... But one could be both selected and current.
   ... [Gives an example with JAWS]

   CS: How does that work for tabs?
   ... Or multiple selections in a listbox?

   BG: With tabs you have aria-expanded.

   JS: You can have multiple tabs expanded, but only one is
   selected.

   CS: I was talking about browser tabs

   BG: [example with remote callback]

   CS: We already have focus and selection.
   ... The difference between focus and selection seems to be the
   difference between selection and current.

   BG: This also applies to static.
   ... People didn't want aria-selected because it's going to
   cause confusion with interactive widgets.
   ... People will start putting it on the wrong widgets.

   CS: Why is that?

   BG: If you can put aria-selected on everything, people are
   going to start putting it on everything

   CS: I think the best mapping in UIA is going to be selected.
   ... I think that's what we do with similar native widgets in
   Windows.

   BG: [Asks about experience in Windows on calendar]

   CS: [Describes it]

   BG: So basically it's the distinction between selected and
   focused is not always the same.

   CS: Correct. Selection and focus are not always the same.

   BG: In the spec, there are many cases where they are the same.

   CS: And in the mapping you'll see that they are all different
   ... Sometimes they are the same, but not always.

   JS: Can I break in to say the meeting is over?

   BG: Can you post this (the Windows mapping) to the list?

   CS: I can post that to the list.
   ... I think I posted something last week, but I can do it
   again.
   ... I think we've gotten as far as we can without further
   input.

   action-1530

   <trackbot> action-1530 -- Cynthia Shelly to Investigate uia
   mapping of role="text" -- due 2014-11-18 -- OPEN

   <trackbot>
   [21]https://www.w3.org/WAI/PF/Group/track/actions/1530

     [21] https://www.w3.org/WAI/PF/Group/track/actions/1530

Summary of Action Items

   [NEW] ACTION: James Craig to investigate the AXAPI mappings for
   aria-current. [recorded in
   [22]http://www.w3.org/2014/11/18-aapi-minutes.html#action01]
   [NEW] ACTION: jcraig to investigate the AXAPI mappings for
   aria-current. [recorded in
   [23]http://www.w3.org/2014/11/18-aapi-minutes.html#action02]

   [End of minutes]

Received on Tuesday, 18 November 2014 21:24:47 UTC