Protocols and Formats Working Group Teleconference
13 Jan 2014


See also: IRC log


joanie, [Mozilla], WuWei, [IPcaller], Jon_Gunderson, Joseph_Scheuhammer, Rich_Schwerdtfeger, Michael_Cooper, Matt_King, janina, marks


<trackbot> Date: 13 January 2014

<richardschwerdtfeger> meeting: W3C WAI-PF ARIA Caucus

<richardschwerdtfeger> be there shortly

<clown> :-)

<scribe> scribe: mattking

HTML 5 defects

<janina> UAIG exclusion (RAND) closed 4 January

<richardschwerdtfeger> https://www.w3.org/Bugs/Public/show_bug.cgi?id=18574

RS: should there be a week mapping between hidden and aria-hidden

<clown> <div hidden aria-hidden="false">This content is available to ATs</div>

RS: That is, should hidden content be available in the accessibility tree in

<SteveF> other related bug https://www.w3.org/Bugs/Public/show_bug.cgi?id=23371

RS: conversely, should aria-hidden without hidden attribute keep content out of accessibility tree
... Microsoft and Apple support weak mapping.

<clown> s/week mappting/weak mapping/

RS: If content is hidden and author puts ari-hidden=false, should the content be exposed to the AT?

David: We have more questions than opinions at this point. If we had very complete spec, we could have more informed opinion.

<SteveF> aria-hidden=false ONLY overrides if set on SAME element as hidden or css display:none

David: Would keyboard behavior be effected?

RS: No, if content is not visible then you can not keyboard operate it.

Joseph: what if you had a button, wouldn' you be able to see the button in the AT and operate it?

Steve: Yes, it would be consistent with off-screen methods.
... old off screen methods do allow you to click or keyboard operate off screen content
... the use case is not for hiding interactive content

RS: html5 guide with mappings is not normative yet.

<SteveF> current UA support for aria-hidden/hidden http://www.html5accessibility.com/tests/hidden2013.html

<janina> UAIG 1.0 implementations call expires Friday 17 January per:

<janina> http://lists.w3.org/Archives/Public/public-pfwg/2013Dec/0028.html

JC: If we do not change aria-hidden to a week mapping to html hidden, then aria-hidden=true does not have any meaning.

David: should we compare to aria-disabled?

RS: aria can not change the functionality of the browser

<clown> <input disabled aria-disabled="false"> ?

JC: IMO, aria-disabled and aria-required could have strong mappings to html 5 disabled and required

<richardschwerdtfeger> https://www.w3.org/Bugs/Public/show_bug.cgi?id=18574

RS: today, ff already has a week mapping in the case when content is visible and aria-hidden is true. The content is visible in accessibility tree and the aria-hidden true is exposed as an object attribute

Joseph: haven't these always had a weak mapping?

JC: It was more like half strong and half week depending the value (true or false)

<SteveF> http://www.html5accessibility.com/tests/hidden2013.html

Steve: All screen readers except Orca hide content where aria-hidden=true even if the content is visually rendered

<joanie> so to finish what I was going to say

<joanie> Orca + WebKit doesn't present aria-hidden:true

<joanie> because webkit doesn't expose it to orca

<joanie> if an AT is not supposed to present something, do not expose it to the AT :)

RS: so we have the week mapping for aria-hidden=true. Now we need agreement on aria-hidden=false.

<Zakim> jcraig, you wanted to state that @hidden doesn't even have a strong mapping to visible display (CSS cascade makes it easy to accidentally show).

RS: asking Alex and David: Do you support using aria-hidden=false to expose visually hidden content to assistive technology?

<Zakim> jcraig, you wanted to state that @hidden doesn't even have a strong mapping to visible display (CSS cascade makes it easy to accidentally show "hidden" elements). <div hidden

<jcraig> [hidden] { display:none; }

<davidb> (there is no "important!")

JC: Currently it is possible for CSS to override html5 hidden so a strong mapping with aria-hidden could cause visible content to not be available in accessibility tree when it should be.

David: I think we are conditionally on board. We have concerns about inheritance and other possible special conditions.

<richardschwerdtfeger> https://www.w3.org/Bugs/Public/show_bug.cgi?id=18574

Steve: can you comment on what you would like to see documented in the bug?

<davidb> there is also https://www.w3.org/Bugs/Public/show_bug.cgi?id=23371

RS: David, can you update bug so we have a record of Mozilla agreement?

David: yes

<davidb> (we should reach out to google as well)

Janina: html5 down to about 10 bugs and we have this one open. We don't want accessibility holding up html5.

Steve: Google is currently supporting the week mapping in android 31.

RESOLUTION bug 18574: solution is accepted to make semantics week and Alex will update bug.

Steve: bug 23371 is the same as 18574 so I have made 23371 blocked. so we do not need separate discussion.


<clown> https://www.w3.org/Bugs/Public/show_bug.cgi?id=23380

<richardschwerdtfeger> https://www.w3.org/Bugs/Public/show_bug.cgi?id=23380

<SteveF> http://www.w3.org/html/wg/drafts/html/master/dom.html#sec-strong-native-semantics

<clown> <input disabled aria-disaled="false">

<clown> ?

Review of strong mappings.

<clown> <input aria-disabled="true">

Joseph: autocomplete should be in this table.

Steve: aria-disabled=true on an html5 element that is not disable

RS: html5 wins so that is an author error

Steve: this bug can be closed because no changes are required in the mappings.
... is that the case? Any objections to closing the bug?

RESOLUTION: close bug 23380

<SteveF> http://www.w3.org/html/wg/drafts/html/master/grouping-content.html#the-p-element

other html 5 discussion

<SteveF> http://www.w3.org/html/wg/drafts/html/master/dom.html#allowed-aria-roles,-states-and-properties

Steve: I am adding allowed aria roles, states, and properties to each html element.

RS: html 5 or 5.1?

Steve: no normative changes; all informative.
... so it could be either html5 or 5.1.
... I am asking for others to review to ehlp us ensure there are not issues with the mappings.

<SteveF> http://www.w3.org/html/wg/drafts/html/master/sections.html#the-article-element

Steve: this can raise other typs of mapping issues that should be reviewed. For example, what should be allowed on BR.

JC: and this could be added to validator

aria 1.0 CR actions and issues

<richardschwerdtfeger> https://www.w3.org/WAI/PF/Group/track/products/1

RS: Michael any progress.

Michael: not last; will happen this week.

RS: new comments?

Michael: none new.

RS: meeting next week?

general agreement, yes.

discussion of next week's agenda

Priority should be 1.0.

<MichaelC> http://www.w3.org/WAI/PF/meetings/2014-01-ftf#agenda

Summary of Action Items

[End of minutes]

Minutes formatted by David Booth's scribe.perl version 1.138 (CVS log)
$Date: 2014/01/13 16:32:19 $

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/

Guessing input format: RRSAgent_Text_Format (score 1.00)

FAILED: s/week mappting/weak mapping/
Succeeded: s/week mapping/weak mapping/
Succeeded: s/aria-disabled/IMO, aria-disabled/
Succeeded: s/these always had a strong mapping/these always had a weak mapping/
Found Scribe: mattking
Inferring ScribeNick: mattking
Default Present: joanie, [Mozilla], WuWei, [IPcaller], Jon_Gunderson, Joseph_Scheuhammer, Rich_Schwerdtfeger, Michael_Cooper, Matt_King, janina, marks
Present: joanie [Mozilla] WuWei [IPcaller] Jon_Gunderson Joseph_Scheuhammer Rich_Schwerdtfeger Michael_Cooper Matt_King janina marks
Regrets: Stefan_Schnabel
Agenda: http://lists.w3.org/Archives/Public/public-pfwg/2014Jan/0006.html
Found Date: 13 Jan 2014
Guessing minutes URL: http://www.w3.org/2014/01/13-pf-minutes.html
People with action items: 

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

[End of scribe.perl diagnostic output]