<trackbot> Date: 07 October 2014

<clown> agenda: this

<scribe> Scribe: Joanmarie_Diggs

<clown> http://www.w3.org/TR/wai-aria/complete#combobox

<clown> http://www.w3.org/TR/wai-aria/complete#implictValueForRole

<richardschwerdtfeger> http://www.w3.org/TR/wai-aria/complete#implicit_semantics

News: ISSUE-541: (All) Change to wording in name/description document re: host language "value" attribute.

<clown> issue-541

<trackbot> issue-541 -- Text Alt Comp too specific to ARIA attributes (aria-valuetext and aria-valuenow) -- open

<trackbot> https://www.w3.org/WAI/PF/Group/track/issues/541

<clown> 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).

JS: There is a step in the accessible name calculation (the old version) where it says something like the above
... The last part was missing until now.

RS: Is it always the value attribute?

JS: I tried to word it so that it didn't actually say "value"

RS: A way to do that would be to make it a glossary item.

JS: Suggested definition?

RS: I can come up with something.
... Text can have a value, but it's not a range.

JS: That's already covered. (Reads text.)

RS: (providing text)

<clown> If the embedded control has role texbox, use its value.

(Group discusses)

RS: The thing about SVG that worries me is that they are making it closer to HTML. So much so that they have a canvas element, iframes.
... We could say for this release that it's limited to HTML.
... The only time I can think that you'd have a value is if you have user input.

JS: I tried to use the word "control" because it is not an HTML element, but captures the fact that it takes input.

(More discussion about what does and does not have a "value")

<richardschwerdtfeger> Value of a control is the is a property of a host languane input control that will take a range of values or text as input

RS: The above is a working glossary definition.

(More working / discussion to refine the language)

<richardschwerdtfeger> Value of a control is a property of a host languane input control that will take a range of values or text as input

<richardschwerdtfeger> http://www.w3.org/html/wg/drafts/html/CR/forms.html#the-select-element

RS: Multiselect is a problem for our definition. The value only corresponds to the first one.

<clown> Value of a control is the is a property of a host languane input control that will take a set of values or text as input

JS: I suggest you propose this as a change to the glossary on Monday.
... I would use set of values instead of range.

RS: I think we could put it in there. Do you want to add that to your proposal?

JS: I guess so.

<clown> issue-541?

<trackbot> issue-541 -- Text Alt Comp too specific to ARIA attributes (aria-valuetext and aria-valuenow) -- open

<trackbot> https://www.w3.org/WAI/PF/Group/track/issues/541

RS: Resubmit it and we can talk about it Monday.

JS: We're talking about issue 541.
... James Craig has a proposal for issue-541 and I was saying that it was already mostly done.

RS: And I'm saying we also need a glossary entry.

JS: There is a "value" in the glossary already.

<clown> "A literal that solidifies the information expressed by a http://www.w3.org/TR/wai-aria/complete#def_state, http://www.w3.org/TR/wai-aria/complete#def_property, http://www.w3.org/TR/wai-aria/complete#def_role, or text content."

<clown> "A literal that solidifies the information expressed by a state, property, role, or text content."

JS: The above is the current definition.

RS: I think we need to change that.

JS: I think one of us should create an issue to change this.

(RS creates an issue)

<clown> issue-541?

<trackbot> issue-541 -- Text Alt Comp too specific to ARIA attributes (aria-valuetext and aria-valuenow) -- open

<trackbot> https://www.w3.org/WAI/PF/Group/track/issues/541

<richardschwerdtfeger> https://www.w3.org/WAI/PF/Group/track/issues/680

ISSUE-542: (All) Use of fall back text attribute for CSS image content rules.

<clown> issue-542?

<trackbot> issue-542 -- Text Alt Comp section on CSS generated content does not allow for proper CSS content fallback -- open

<trackbot> https://www.w3.org/WAI/PF/Group/track/issues/542

JS: This is something we discussed yesterday. Another suggestion from James about accessible name calculation.

<clown> <div style="content: url(icon.png), attr(data-fallback);" data-fallback="Bar">Foo</div>

JS: The above is his example.
... You can now in CSS add an attribute.
... This provides a data fallback as your textual content.
... So the alternate name is Bar.
... James is suggesting that because you can do this in CSS, we need to add text that references this technique.
... I gather that the attribute can be anything you want it to be (e.g. aria-label)

RS: What does it do?

JS: That style command says put this icon image here.

<clown> !

JS: The attribute says "use the following attribute for the text alternative" on that element

<clown> <div style="content: url(icon.png), attr(aria-label);" aria-label="Bar">Foo</div>

JS: But I think you could also write something like the above.
... But this isn't quite right as you have to use the before and after stuff.

RS: So you can use an ARIA label instead of an alt?

JS: You can't use alt because it's not an attribute of div
... It has to be a supported attribute, as I understand it from yesterday's discussion.

RS: The thing is, I just don't know how wide open this content is going to be. Can it be anything?

JS: You'd have to look at the CSS spec as I don't know it that way.
... The other thing I recall from yesterday is that only Safari implements it so far.

<clown> issue-542?

<trackbot> issue-542 -- Text Alt Comp section on CSS generated content does not allow for proper CSS content fallback -- open

<trackbot> https://www.w3.org/WAI/PF/Group/track/issues/542

JS: Does this look familiar to you Cynthia from IE's perspective?

CS: I'm not familiar with this feature.
... I'll see if I can find more information on this.

(Group members individually looking at the CSS spec)

<richardschwerdtfeger> http://www.w3.org/TR/CSS21/generate.html

RS: (Reads from above spec)
... The issue is that now CSS is injecting new forms of content.

CS: If this is in the spec, then, yes, this is indeed an issue.

JS: It's shipped and Safari reportedly implements it.

<clown> "This function returns as a string the value of attribute X for the subject of the selector. The string is not parsed by the CSS processor. If the subject of the selector does not have an attribute X, an empty string is returned. The case-sensitivity of attribute names depends on the document language."

RS: So, yes, you could put an aria label in there.

JS: Yes anything that is a valid attribute

CS: We should get WCAG in on this.

(Discussion about computed style)

<clown> $.css ('content');

JS: $ stands for the element in question. It comes back as a string.

<clown> "url(icon.png), attr(aria-label)"

RS: Which you now have to parse. That's ugly.

CS: I don't really like CSS-generated content.
... James is right, it's an issue. We need to account for it.

JS: So as far as the name computation is concerned, I have to put something about this.

RS: An entire list of attributes, and you have to be aware of them and see if any of them result in text that can be used. Right?

JS: That's a very big list.

RS: Exactly

CS: Do we have time at TPAC with CSS? We should work with them on this.

RS: I will email Janina and copy the group and suggest we should discuss this because it poses a significant challenge for testing tools.

<clown> <img style="content: url(icon.png), attr(src);" src="Bar">

<clown> <img style="content: url(icon.png), attr(src);" src="./images/foo.png">

CS: I think there's also a WCAG technique here.

RS: Are you in WCAG?

CS: Not for a while.

RS: I'll send the email.

JS: Next Monday is Thanksgiving in Canada.

(Discussion about holidays next week)

JS: When is the discussion about name computation at TPAC?

RS: I don't think we have an exact time yet.

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)

No ScribeNick specified.  Guessing ScribeNick: joanie
Found Scribe: Joanmarie_Diggs
Default Present: Joseph_Scheuhammer, Joanmarie_Diggs, Bryan_Garaventa, Rich_Schwerdtfeger, [Microsoft]

WARNING: Replacing previous Present list. (Old list: Cynthia_Shelly)
Use 'Present+ ... ' if you meant to add people without replacing the list,
such as: <dbooth> Present+ Joseph_Scheuhammer, Joanmarie_Diggs, Bryan_Garaventa, Rich_Schwerdtfeger

Present: Joseph_Scheuhammer Joanmarie_Diggs Bryan_Garaventa Rich_Schwerdtfeger Cynthia_Shelly

WARNING: No meeting chair found!
You should specify the meeting chair like this:
<dbooth> Chair: dbooth

Found Date: 07 Oct 2014
Guessing minutes URL: http://www.w3.org/2014/10/07-aapi-minutes.html
People with action items: 

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