W3C

Protocols and Formats Working Group Teleconference

10 Nov 2015

See also: IRC log

Attendees

Present
Joanmarie_Diggs, Rich_Schwerdtfeger, Bryan_Garaventa, Joseph_Scheuhammer, Cynthia_Shelly
Regrets
Chair
Joseph_Scheuhammer
Scribe
Joanie

Contents


<trackbot> Date: 10 November 2015

<clown> agenda: this

<clown> topic WebEx Access Code 641 707 405 (Next meeting 17–Nov-2015)

<richardschwerdtfeger> I have to leave after 25 minutes - just a heads up

<scribe> scribe: Joanie

Reminder: check that all new ARIA 1.1 spec have actions for the Core-AAM.

JS: I wanted to remind everyone that if you see something in the spec which needs to be dealt with in the core mapping, please let me know.
... So we can create actions.

ISSUE-723 - 734: Updated UIA role mappings (mostly landmarks). Close these issues?

JS: This requires Cynthia.

<clown> issue-723?

<trackbot> issue-723 -- UIA mapping for application incorrect -- raised

<trackbot> https://www.w3.org/WAI/PF/Group/track/issues/723

JS: Cynthia filed just over 10 issues about certain UIA mappings.
... For example issue-723 is the mapping for role application.
... (Reads corrected mapping)
... While you were at TPAC, I've updated these mappings as per Cynthia's issues.
... Many of these are landmarks.
... There is an issue with search.

<clown> http://w3c.github.io/aria/core-aam/core-aam.html#role-map-search

JS: I had to update the UIA mapping, and I noticed there is nothing for MSAA+IA2.
... Then I looked at the release and there's nothing there either.
... We've never had a mapping for this, going back to 2008.
... Does anyone know what it is?

BG: No.

JS: The cell is just blank and has been blank for a long time.
... Can I give someone an action to figure out what it is?

<clown> ACTION: Joanie to determine what the MSAA+IA2 mapping for role="search" [recorded in http://www.w3.org/2015/11/10-aapi-irc]

<trackbot> Created ACTION-1738 - Determine what the msaa+ia2 mapping for role="search" [on Joanmarie Diggs - due 2015-11-17].

JD: If you mean IA2, I have recently set up a Windows box and can look.

JS: Action crated.
... What were you planning to use?

JD: AccProbe

RS: Chrome and Firefox each use IA2.

DPub Mappings

RS: I would like people to look at the DPub mappings if they have the time.

<richardschwerdtfeger> http://rawgit.com/w3c/aria/master/dpub-aam/dpub-aam.html

RS: This is going to be a FPWD. (link above)

JS: At this point, they just have roles, right?

RS: Correct.
... We've listed a number of issues in here.
... I'd like to see if anyone finds any significant issues/concerns.
... They have a number of links in here. I don't know if they'll stay. And there is an Editor's note about it. So I wouldn't be too concerned about those.
... I also have some TBDs in there, if anyone has any details to fill them in.

JS: How did navigation fare?
... I set up a couple of tables for Tzviya because she wanted to see how the tables were done.
... In one of them, I said it was done the same way navigation was done.
... I noticed that in many cases the mapping for IA2 and ATK are object attributes.

JD: We have quite a few roles in ATK which will map, so I'll take a look.

<clown> <table role="navigation" …>

ACTION-1527: (Joseph) UIA mapping of aria-current. Review edits.

<clown> action-1527

<trackbot> action-1527 -- Joseph Scheuhammer to Investigate the proper uia mappings for aria-current -- due 2015-10-01 -- OPEN

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

JS: I've put in the edits Cynthia suggested for UIA.

<clown> http://rawgit.com/w3c/aria/master/core-aam/core-aam.html#ariaCurrent

JS: Whatever the value of aria-current is, is exposed. It's a name-value pair.
... I know that sometime in the future this may change, but she indicated in the action that for now this mapping is what should be done.

RS: It's a name-value pair where?

JS: In the AriaProperties.
... You can think of AriaProperties as ATK's object attributes.

Cynthia: issues with text alternative calculation.

<clown> action-1737?

<trackbot> action-1737 -- Joseph Scheuhammer to Determine mappings for ARIA 1.1 table and cell role for all AAPI platforms. -- due 2015-11-13 -- OPEN

<trackbot> https://www.w3.org/WAI/PF/Group/track/actions/1737

ACTION-1737: (Joseph) Mappings for table and cell roles. Review edits.

JS: I did the mappings for table and cell roles for all platforms.

<clown> http://rawgit.com/w3c/aria/master/core-aam/core-aam.html#role-map-table

JS: The above is table.

JD: The role mapping is correct, but you missed the ATK/AT-SPI2 table interface.

JS: The directive was to copy what's in grid, but change the xml-role.

JD: Grid should also implement the table interface on my platform.

<clown> http://rawgit.com/w3c/aria/master/core-aam/core-aam.html#role-map-table

JS: Is the UIA mapping correct, Cynthia?

<cyns> https://msdn.microsoft.com/en-us/library/windows/desktop/ee671650(v=vs.85).aspx

JD: Oh wow, no interfaces at all in the Core mapping spec for my platform.

action Joanie to provide all the missing ATK/AT-SPI2 interfaces for Joseph

<trackbot> Created ACTION-1739 - Provide all the missing atk/at-spi2 interfaces for joseph [on Joanmarie Diggs - due 2015-11-17].

CS: The control type is table; not data grid.

<cyns> Control Type = Table

<cyns> which implements the Grid and Table Patterns

CS: I also put a link to the documentation in (above)

JS: Both Grid and Table patterns?

CS: Yes. A Table is a Grid with headers.

JS: I believe that I got the AX API mappings from Joanie.

JD: I'm not remembering that, but perhaps.... :)
... I think it was an earlier meeting, and someone who was not scribing had a Mac handy.

JS: Moving on to cell

<clown> http://rawgit.com/w3c/aria/master/core-aam/core-aam.html#role-map-cell

JS: The role mappings are above.

CS: So this is a table cell; not a grid cell?

JS: Correct.

<cyns> Control Type = TableItem

JD: I'm pretty sure that IA2 has a table-cell interface because we "stole" it for ATK/AT-SPI2.

<cyns> Must support selection pattern and must not support invoke pattern

<cyns> https://msdn.microsoft.com/en-us/library/windows/desktop/ee671649(v=vs.85).aspx

JS: Cynthia, should table also support selection?

CS: No. Table pattern and Grid pattern for ARIA role table.
... I don't think our table supports invoke, so I don't think that will be a problem.
... We just made a change that anything with an onclick handler supports invoke.

JS: Other questions or comments on cell?

Reminder: check that all new ARIA 1.1 spec have actions for the Core-AAM.

ISSUE-723 - 734: Updated UIA role mappings (mostly landmarks). Close these issues?

JS: Cynthia, you filed a number of issues which were to change the UIA role mappings for a number of things.
... I have made the changes. Do you want to confirm them?

CS: Yes.

<clown> http://rawgit.com/w3c/aria/master/core-aam/core-aam.html#role-map-application

<clown> http://rawgit.com/w3c/aria/master/core-aam/core-aam.html#mapping_role_table

JS: You can go to this URL (above). You'll see a button that says "view as a single table".
... Then you can show/hide columns, so you can hide everything that's not UIA.

CS: Banner looks good.
... Complementary looks good.
... Contentinfo should be content information
... Main is correct.
... Search is correct.
... Navigation is correct.
... Article... I may have misunderstood the purpose of this role after having spoke with Matt King. So let's leave the article issue open for now.

JS: I don't think Article is a landmark.

<clown> issue-730

<trackbot> issue-730 -- UIA mapping for article incorrect -- raised

<trackbot> https://www.w3.org/WAI/PF/Group/track/issues/730

CS: Definition is correct.
... Log is correct.
... Math is correct.
... Note is correct.

JS: I will close all the issues but application and article.
... I'll also close content info after I have correct it.

CS: Sounds good.

ACTION-1527: (Joseph) UIA mapping of aria-current. Review edits.

<clown> http://rawgit.com/w3c/aria/master/core-aam/core-aam.html#ariaCurrent

JS: You gave me the answer Cynthia. I just want to make sure I got it right.
... (reads it)

CS: That is correct for now.
... The link you gave me says "not mapped"
... Correction, wrong one.

RESOLUTION: Close action-1527.

<clown> +1

<clown> close action-1527 based on consensus from the 10-Nov-2015 AAPI meeting.

close action-1527

<trackbot> Closed action-1527.

note action-1527 Closing based on consensus from the 10-Nov-2015 AAPI meeting.

<trackbot> Notes added to action-1527 Investigate the proper uia mappings for aria-current.

Text Alternative Computation

CS: I think I'll put in issues for things I've found.
... We started trying to implement it and read into issues due to vagueness in the text.

JS: As far as I know, it's the same as 1.0. Just editorial.

CS: There was a high-level issue trying to piece together the ARIA and HTML computations.
... I know we intentionally separarted it. But when you go to implement it, it's really confusing.
... I spent multiple hours working with my dev.
... The other high-level piece of information is that that ARIA one is much more complete than the HTML one.

JS: I get the impression that the intended audience for the ARIA one is authors.
... And the intended audience for HTML, implementors.
... If you create the issues, Cynthia, we will then have open items we can discuss.

<clown> http://rawgit.com/w3c/aria/master/accname-aam/accname-aam.html

<clown> http://rawgit.com/w3c/aria/master/accname-aam/accname-aam.html#mapping_additional_nd_te

CS: The first comment is on step 2a.
... (Reads step 2a)
... We only care if we reach the current node via computation; not if it's contained in any reference (gives example).

JD: You have an example you can share?

CS: I'll put it in the issue.
... (Points out other issues fairly rapidly, but tells scribe she will put them in the issues she creates.)
... There are several places where the ARIA one mentions the host-language one, but the HTML one doesn't address it.

<clown> http://www.w3.org/TR/2014/REC-wai-aria-implementation-20140320/#mapping_additional_nd

JS: (Reads note from the 1.0 version for which the link is above)

CS: It was unclear what HTML elements and attributes applied.
... Do we own that?

JS: HTML AAM is part of the ARIA github tree.
... The editors are Steve, Alex, and Jason.
... I would like them to replicate the structure of the SVG AAM.

CS: I think I'll start with issues here and then figure out which need to be moved into bugs against them.
... And we can point to the SVG AAM if and as appropriate.

JS: We can do that.

CS: Are they using bugzilla or github or...?

JS: Bugzilla.

Summary of Action Items

[NEW] ACTION: Joanie to determine what the MSAA+IA2 mapping for role="search" [recorded in http://www.w3.org/2015/11/10-aapi-irc]
 
[End of minutes]

Minutes formatted by David Booth's scribe.perl version 1.140 (CVS log)
$Date: 2015/11/10 22:10:48 $