RE: DRAFT MINUTES: October 6, 2014 WAI-PF ARIA Caucus

Belated regrets.  I tried to call in, but the bridge was full.

From: Gunderson, Jon R [mailto:jongund@illinois.edu]
Sent: Monday, October 6, 2014 11:37 AM
To: W3C WAI Protocols & Formats
Subject: DRAFT MINUTES: October 6, 2014 WAI-PF ARIA Caucus

Draft Minutes:

http://www.w3.org/2014/10/06-aria-minutes.html

[W3C]<http://www.w3.org/>
- DRAFT -
Protocols and Formats Working Group Teleconference
06 Oct 2014
See also: IRC log<http://www.w3.org/2014/10/06-aria-irc>
Attendees
Present
Regrets
Chair
Rich
Scribe
jongund
Contents

  *   Topics<http://www.w3.org/2014/10/06-aria-minutes.html#agenda>
     *   TPAC<http://www.w3.org/2014/10/06-aria-minutes.html#item01>
     *   Name computation (issue 541)<http://www.w3.org/2014/10/06-aria-minutes.html#item02>
     *   Issue 542<http://www.w3.org/2014/10/06-aria-minutes.html#item03>
  *   Summary of Action Items<http://www.w3.org/2014/10/06-aria-minutes.html#ActionSummary>

________________________________
<trackbot> Date: 06 October 2014
<richardschwerdtfeger> meeting: W3C WAI-PF ARIA Caucus
<richardschwerdtfeger> http://lists.w3.org/Archives/Public/public-pfwg/2014Oct/0008.html
<richardschwerdtfeger> http://lists.w3.org/Archives/Public/public-pfwg/2014Oct/0008.html
TPAC
ready
Janina: We almost have a calendar, thursday is booked
... Starting with WebTV, best time
... 11:15 HTML working group on two topics
... Accessible graphics and event handler enumeration
JC: Recommend the PF group only raise this issue again if a technical proposal is ready to go with it, we need at least one implementor on board in order to raise this again
Janina: We need a separate time to discuss this recommendation
... We are concerned about DRM stuff getting in the way
RS: Comcast is building an accessible set top box
Janina: I have also seen it at CSUN
RS: Is there something specific?
... What are they looking to standardize?
Janina: Say something that long, so there are some opportunities that would have captions and multiple languages, can be used for testing, talking to mark about this
RS: Isnt there a mandate
Janina: Not all in a mandate, especially in audio descriptions
RS: You have to have something concrete
JC: The same thing probably applies, event enumeration has be raised and declined many times
Janina: We have a chain of e-mails on the list
JC: The original reasons were not on the list
Janina: Shane stated security was one of the main reasons
JN: Tools can already do this, they just need to write plateform specific test tools
JC: I agree
RS: We are going to have UI events, and design patterns, I am more focused on where we are going, than where we have been
JN: It is the same issue on security
RS: Query event handlers?
JC: Everything we need is already there, this group has already rejected this
Janina: We need to put hat together on the compelling reason
JN: We should drop it
JC: One of my directors sometimes says, "I only have a few silver bullets... I don't want to waste of them one on this issue."
<Zakim> Joseph_Scheuhammer, you wanted to discuss longdesc
RS: Please cannot discuss longdesc?
Janina: Can we go back to the agenda
... We are still working on nailing the meeting done, 66% done
... Digital publishing in the after lunch
<richardschwerdtfeger> +1
Janina: We have another 30 minutes with the HTML group
... We then have time with Mark H
RS: The big things are the taxonomy for graphics, and the validation issues
... the whole think with keyboard, the biggest advances are CSS and web apps
... It would be nice to have a modal dialog box, but we don't see that much on mobile
JC: Another issue is data providers for tables, everyone has there own
... It is a huge issue for accessibility and performance
... I am not sure I have time to prepare a summary of the topic
Janina: So not something they should work on in the near future
JC: I think we can ask them too
... Let me think about it more, there are major benefits for accessibility and some on performance, it would take me some time to think about it
Janina: We want to think about the future and take advantage of meeting time with other groups
JC: The way most table views, similar to set-size and ste-position, maybe only a couple 100 view elements, some unordered view
... only the element in the current view are available, the data source will provide more when needed, this is all managed by the author now
MK: How does pos inset, press the down arrow, it requests a row element and gets the next row
JC: You can easily get ahead of what the data is available, so there are gaps in the user view of information
... Let me search for something in the table, but not all the data is there in these rendered views
... milliseconds versus double-digit seconds
... Scroll wheel scrolling has similar effects, an API to get to these data sources is much more effecient
MK: If there is something not rendered, I would not expect a screen reader to search it
JC: There maybe data in the memory, but the data is not rendered
... if you had something like safari native find it could search these APIs
MK: How you want to do that search is different between applications
JC: All native platforms render table views and list views this way, but the Web doesn't have a solution for table and collection view data providers.
MK: I don't see how solving it for accessibility is only for accessibility
JC: It is not specific for accesibility, but it will help
RS: Anything else on this one?
Jania: Not for me
RS: The DOM thing would be nice, but the other ones are critical
MK: There was a discussion about parametrized queries?
RS: That wil move to 2.x
Name computation (issue 541)
<clown> issue-541
<trackbot> issue-541 -- Text Alt Comp too specific to ARIA attributes (aria-valuetext and aria-valuenow) -- raised
<trackbot> https://www.w3.org/WAI/PF/Group/track/issues/541
<richardschwerdtfeger> https://www.w3.org/WAI/PF/Group/track/issues/541
<clown> http://rawgit.com/w3c/aria/master/accessiblename/accName.html#mapping_additional_nd_te
JC: I have proposed text there
<jcraig> Current text: If the embedded control is a range (e.g., a spinbutton or slider), use the value of the aria-valuetext attribute if available, or otherwise the value of the aria-valuenow attribute.
<jcraig> Proposed: If the embedded control is a range (e.g., a spinbutton or slider), use the text value if defined (e.g. aria-valuetext), or otherwise the value of the control (e.g., aria-valuenow, or host language value attribute).
JC: This is editroial
RS: I agrre
... Can we assign this to the implementation guide
JS: We are close tot being done
RS: Put it on the 1.1 implementation guide
MC: We can do it on the implementation guide or something new
RS: JS are you all set?
JS: When I reviewed all the HTML references it was language similar to JC wording, I like JC wording
<richardschwerdtfeger> https://www.w3.org/WAI/PF/Group/track/issues/542
Issue 542
RS: Proper fall back
JC: This has slipped between the cracks in the CSS spec
... If its using comma separated lists for alternative content, webkit supports, I doubt there is other implementation
... There is an ALT property in the CSS specs
... The test tools need to look at the live DOM and this information is the DOM
JS: Is that specific to HTML or can it also be used in SVG
JC: The data-* attributes are specific to HTML, but the attr() content is not.
JS: It can be attribute X
JC: ALT doesn't mean anything on divs
... It is for localization via the stylesheet
JS: We have COR, HTML specific and SVG specific, where does this go?
JC: Let me change the example, you can't render content directly on the element
RS: If this is going into CSS, we need to look at it
<jcraig> <div data-fallback="Bar">Foo</div>
<jcraig> div::after {
<jcraig> content: url(icon.png), attr(data-fallback);"
<jcraig> }
JC: This has been in since CSS 2
RS: Do you sit in on all this stuff
JC: I don't go to the calls, but I follow the specs
RS: I have not seen people use it
JC: It works in one browser
... Chicken and egg problem
Janina: If we talk about this friday morning, who should we invite from CSS
RS: We need to look at this alot more
<Birkir_> Birkir Gunnarsson on phone too, long time absent. 919 number.
JC: This has only been added recently
... fall back text content ....
RS: Are there more things we need to look at
JC: Potentially....
RS: The separation of content and presentation is getting blurry, so this is a concern
JN: Which CSS spec is this in?
JC: CSS 4 selectors...
JN: I will take a look
RS: Is WCAG looking at this? This is a big can of worms, test tools and authoring practices...
ISSUE 542
<MichaelC> issue-542?
<trackbot> issue-542 -- Text Alt Comp section on CSS generated content does not allow for proper CSS content fallback -- raised
<trackbot> https://www.w3.org/WAI/PF/Group/track/issues/542
MC: What are you asking me?
RS: The issue here, you have the ability to add ALT content form CSS instead of the HTML source
... It has to go into the name commputation, authoring practices, we need to have a closer look at this CSS feature, we need to pay ...
<Birkir_> assistive technologies, at least screen readers, do not search for alt content of CSS either (at lesat speaking definitivel for Jaws/NVDA)
JC: We discussed this, but didn't get proposed to the CSS working group
... It was this group's inaction that made me propose to the CSS working group separately, because it was delaying accessibility of several Apple products.
RS: Other that JC, we are not looking at this holistically
Janina: I think WCAG will pay attention to it, we also need to pay more attention to it
RS: Can we create an action for the WCAG group look at this
MC: There are several people in this group that are in the WCAG group, we can do it as a chair to chair thing, or as a technique group issue
... We still have to get someone to write a technique, and then we need to make sure there is accessibility suppoort
... We need to wai for implementation and create test cases
RS: Can we get access to this through the DOM in webkit?
<MichaelC> qq-
JC: In webkit nightly we make label available in the DOM and it will include CSS information
... 5 or 6 years in webkit
RS: Current tools are using the DOM, is there a way to get at it in the CSS ....
JS: It might not be able to get the computer label, so you want a way to get the computed label
<jcraig> CSS "alt" discussion on www-style: http://lists.w3.org/Archives/Public/www-style/2012Nov/0319.html
MC: You may mean test the WCAG techniques, but the group recommends testing for more ....
<Zakim> MichaelC, you wanted to clarify ¨testing to WCAG compliance¨
MC: We need to have accessibility support, even if it can in theory be accessible
... There is a CSS object model spec, but I don't know alot about it
<MichaelC> CSS Object Model<http://www.w3.org/TR/cssom/>
RS: CSS is becoming more popular that Javascript
Janina: Can we get some time at TPAC on the topic
MC: I recollect that there was some work, but it may have been dropped
<jamesn> jcraig, I saw the mail. I still can't find it in a spec. I'm concerned about how this works. Calling it fallback seems incorrect and I just want to see how it is specified
RS: RS: You mean an accessibility API mapping?
... We had talked about a accessible name mapping
JC: That was proposed for the DOM.....
... get computed role and get computed ... I discussed in January
RS: One way or another we need to get this information, either the final name and or through the CSS
JN: This can be done in a browser specific environment, through extensions
... Simiar to even model stuff
RS: Is there a way to do this across browser
JN: If you are using Firefox and Webkit you have access to this information through extensions
RS: It depends on how much they are doing with CSS
JG: We do this with OpenAjax tools
RS: I am concerned that CSS is changing features on how we find content in the browser
... Only in the last week we have been looking at this from a tool perspective
JN: I know about generated content; I wrote the tool we use as a FF extension
JC: I wrote one for webkit
<clown> https://www.w3.org/WAI/PF/testharness/testresults?testsuite_id=1&testcase_id=553
MK: We used generated content in place of sprits
JS: There is a test case in the test suite test for this
RS: We don't have it for this one, for generating ALT text
... We need to look at this in much more detail
Janina: We will look at this all Friday morning
RS: JN if someone puts in an image VIA CSS
JN: FF doesn''t support it yet, but when it does it will
... It is required to have a text alternative
RS: The group meeting I was in required to have ALT attribute
JN: There was a big discussion about this in WCAG, there are in the techniques now, we did not get TITLE
RS: We beat that to death, ....
JN: There is not technique for TITLE
<Zakim> jcraig, you wanted to talk about CSS WG process
JC: We need more people in other working groups with accessibility interests...
<MichaelC> WCAG-ACTION-282?
<trackbot> WCAG-ACTION-282 -- Michael Cooper to Raise issue of css technique for generated content fallbacks https://www.w3.org/wai/pf/group/track/issues/542 -- due 2014-10-13 -- OPEN
<trackbot> http://www.w3.org/WAI/GL/track/actions/282
Janina: It has to be a two way street
... We did that in HTML reasonably well
JC: Rich is in SVG
RS: We need someone in CSS
JC: We need more people in CSS with accessibility concerns
RS: we need to do that
... like the canvas thing, we need to have someone to work in that area
<richardschwerdtfeger> https://www.w3.org/WAI/PF/Group/track/issues/609
RS: Any proposed text?
JC: There is proposed text, so assign to implementation guide
RS: Signed over
ISSUE-642
<trackbot> ISSUE-642 -- alt text computation should distinguish between nameFrom contents during root node calculation versus recursive calculation. -- raised
<trackbot> https://www.w3.org/WAI/PF/Group/track/issues/642
<richardschwerdtfeger> https://www.w3.org/WAI/PF/Group/track/issues/642
RS: We don't do that
JC: JS just commented, so this is a larger name that needs to happen
RS: JS can you take this on?
JS: Sure
JC: This was a constant point of confusion in the psuedo code
RS: A section on name and description
Janina: Do we want to talk to the HTML time?
JC: We should invite SF to the discussion
... Jason Kiss too
<clown> action-1073?
<trackbot> action-1073 -- Matthew King to Update aria-selected to reflect that it communicates selectability and clarify responsibility for ensuring aria-selected=false is on selectable elements -- due 2014-09-29 -- OPEN
<trackbot> https://www.w3.org/WAI/PF/Group/track/actions/1073
<richardschwerdtfeger> https://www.w3.org/WAI/PF/Group/track/actions/1073
MK: Some how I got this, I did not ask for
... I am still willing to work on it
... Lets put it out three weeks
RS: After TPAC, how about November 18th
ACTION-1335
<trackbot> ACTION-1335 -- Joanmarie Diggs to Propose edit for issue-493 to explicitly disallow strings matching "" or "undefined" in this sentence: content authors must provide values for required states and properties. -- due 2014-09-30 -- OPEN
<trackbot> https://www.w3.org/WAI/PF/Group/track/actions/1335
<clown> action-1335
<trackbot> action-1335 -- Joanmarie Diggs to Propose edit for issue-493 to explicitly disallow strings matching "" or "undefined" in this sentence: content authors must provide values for required states and properties. -- due 2014-09-30 -- OPEN
<trackbot> https://www.w3.org/WAI/PF/Group/track/actions/1335
Joanie: I am learning how to edit the spec tomorrow, can we push it back a week
RS: You want to put a "author must"
... We will have to test
<clown> http://www.w3.org/TR/wai-aria/complete#requiredState
JC: We just say does this content pass it ...
RS: We do have to cover it with the director
JS: it is already in the spec...
RS: Just a clarification, are you happy with the text
JC: I said my language is clunky, but it is accurate
RS: A week
JC: You are already familiar with git?
<jcraig> Content authors must provide a non-empty value for required attributes; this value may not match the format "undefined".
Joanie: Yes
JC: ... discussing spec editing and who should do what and outcomes ....
Summary of Action Items
[End of minutes]
________________________________
Minutes formatted by David Booth's scribe.perl<http://dev.w3.org/cvsweb/%7Echeckout%7E/2002/scribe/scribedoc.htm> version 1.138 (CVS log<http://dev.w3.org/cvsweb/2002/scribe/>)
$Date: 2014-10-06 18:29:41 $
________________________________
Scribe.perl diagnostic output
[Delete this section before finalizing the minutes.]
This is scribe.perl Revision: 1.138  of Date: 2013-04-25 13:59:11
Check for newer version at http://dev.w3.org/cvsweb/~checkout~/2002/scribe/<http://dev.w3.org/cvsweb/%7Echeckout%7E/2002/scribe/>

Guessing input format: RRSAgent_Text_Format (score 1.00)

Succeeded: s/at least one browser implementing/at least one implementor on board in order to raise this again/
Succeeded: s/Recommend that a solution in place, it has come up and been shot down/Recommend the PF group only raise this issue again if a technical proposal is ready to go with it/
Succeeded: s/My director says I only have silver bullets, is this one we want to use a bullet it on?/One of my directors sometimes says, "I only have a few silver bullets... I don't want to waste of them one on this issue."/
Succeeded: s/milliseconds versus seconds/milliseconds versus double-digit seconds/
Succeeded: s/Other plateforms use APIs, but the web doesn't have a solution/All native platforms render table views and list views this, but the Web doesn't have a solution for table and collection view data providers./
Succeeded: s/All native platforms render table views and list views this, /All native platforms render table views and list views this way, /
Succeeded: s/workings/wording/
Succeeded: s/I think right now it is specific to HTML/The data-* attributes are specific to HTML, but the attr() content is not./
Succeeded: s/It was this groups inaction that made me proposed to the CSS working group/It was this group's inaction that made me propose to the CSS working group separately, because it was delaying accessibility of several Apple products./
Succeeded: s/mush/much/
Succeeded: s/Firefox and Webkit you have access to this information/Firefox and Webkit you have access to this information through extensions/
Succeeded: s/I know about generated content/I know about generated content; I wrote the tool we use as a FF extension/
Succeeded: s/GET/git/
No ScribeNick specified.  Guessing ScribeNick: jongund
Inferring Scribes: jongund

WARNING: No "Present: ... " found!
Possibly Present: Apple Birkir_ Bryan_Garaventa Fred_Esch GVoice IPcaller JC JG JN JS James_Nurthen Jania Joanie Joanmarie_Diggs Jon_Gunderson Joseph_Scheuhammer LJWatson MC MK Matt_King MichaelC Michael_Cooper P10 P14 Proposed RS Rich_Schwerdtfeger Stefan Susann_Keohane aaaa aria bgaraventa1979 clown content div fesch https jamesn janina jcraig joined mattking richardschwerdtfeger trackbot
You can indicate people for the Present list like this:
        <dbooth> Present: dbooth jonathan mary
        <dbooth> Present+ amy

Found Date: 06 Oct 2014
Guessing minutes URL: http://www.w3.org/2014/10/06-aria-minutes.html
People with action items:

WARNING: Input appears to use implicit continuation lines.
You may need the "-implicitContinuations" option.


[End of scribe.perl<http://dev.w3.org/cvsweb/%7Echeckout%7E/2002/scribe/scribedoc.htm> diagnostic output]