Protocols and Formats Working Group Teleconference -- 02 Jun 2014

<trackbot> Date: 02 June 2014

<richardschwerdtfeger> meeting: W3C WAI-PF ARIA Caucus

<richardschwerdtfeger> meeting: make log public

<richardschwerdtfeger> http://lists.w3.org/Archives/Public/public-pfwg/2014Jun/0001.html

zakiim scribe, jongund

RS: Do we need more test cases for HTML5
... More for ARIA 1.1
... We have native host language semantics in ARIA 1.1
... TR has the same as role=row

MC: We were talking about this earlier
... the strong native sematics needs testing, if not implemented it would need to be removed

RS: Fundamental difference

MC: Does the HTML working group need to testing

RS: If they say will not be able to over right ....
... I think we need additional test cases or take it out

MC: We did accommodate, as they progress they will need to test

RS: As the host language has the native semantics we may need to test

MC: We will need to test

RS: They don't have such a document, do they need to test

JS: They seem have a permissive approach to testing, they test some things and not others
... they need to chow implementatino

RS: We are in better shape for 5.1

JS: they will say it is an edge case
... they are trying to get to last call, met with TBL last week
... TBL may be asking about accessibility
... We are OK with most stuff in 5.0

RS: RS: The need an API, almost like it is not testable since no mapping guide

MC: I think it is possible to test, without guide it is more difficult
... It will be better with the mapping guide

RS: You could a validator for testing that stuff?
... What will HTML validator do for conflicting information

MC: You can test it

JS: What will happen in the accessibility API

<Zakim> Joseph_Scheuhammer, you wanted to ask: what is this, if not a mapping guide:

RS: Better question, but they don't have a mapping guide

MC: If it is a user agent requirement, using the validator helps identify that is a condition, validator probably cannot be used for CR requirements

RS: I beleive thats the best they can do

JS: Do we prefer the section be removed?

RS: I care about ARIA 1.1

MC: I don't have a preference
... they made a big deal to add it

JS: Here is the problem, then we need to renegociate, if its there it will be there in 5.1
... If we like it we develop the mapping guide and file bugs

RS: they could make the session non-normative
... They could say "This could be normative in a future spec"

MC: YI think that is viable

RS: We could have them test it
... aria-checked would trump checked in FF

Clown: I agree that it is a possibility, not consistent

RS: make it non-normative

MC: We say that the section needs testing, it is expected to have problems
... Do we say up front that it should be non-nromative or wait until they have problems

RS: We should probalbt say it up front it is not testable

<clown> http://www.w3.org/html/wg/drafts/html/master/dom.html#wai-aria

MC: We are talking about strong native semantics

RS: I don't know how they test without an API

Clown: they need to create test files and test them on 4 platforms

MC: This was tested in our ARIA testing
... Give them a couple of tick marks

JS: i want to keep our working relationships, and so be straight forward with them

MC: Say some sections are not normative

JS: That will change as we develop the mapping guide

MC: We have consensus for HTML

JS: Joseph can you take this up on the tuesday call

Clown: That is not in code?

JS: You should start calling it core
... You can decide how often to talk about the HTML mapping
... On the HTML side we need to editors which tree it should be in
... need to ask the editors
... I have what we need
... We are basically happy with 5.0, we do have a few problems, we just talked abotu some of them

RS: I want to keep the relationship good, like with the canvas work

JS: We will not have everything in the 5.0 time frame

RS: OK
... What does that do for the rest?

MC: I don't think it does

RS: Do they define role computation?

MC: no

RS: we are good
... Is chris G on today?

MC: I would liek to talk about publishing, before other topics

RS: Righ the heart beat draft

MC: I have been working on publication, I am not sure it is ready to go and we need a resolution to publish

RS: I had some problem with the status section

MC: That's because it is old content, i will be updating it
... I updatied with editors draft

RS: I think they need to fix 5.5.3

Clown: I see it in the list

RS: OK, good
... Abstract stuff, I think you saw that, RDF stuff is still there

MC: I don't think it needs to be in the spec, some people might be using it

RS: We need to link to HTML5, not HTML 4

JS: I agree
... Doesn't sound like publication this week

MC: Need to sort out who will make the edits
... The user agent implementation guide needs some more work

Clown: Not called that anymore

MC: thats one of the issues

RS: Mark as a defect and fix in next release

MC: I think it can wait, but we hope to get public feedback

JS: I think we should drop the RDF

Clown: What does that mean?

MC: Dropping the appendix

<MichaelC> https://rawgit.com/w3c/aria/master/spec/aria.html

RS: not quite
... Let me find you an example
... We have related concepts ...., a whole section on RDF in here
... We have RDF properties, in section 5.1

MC: That section might not be ...

RS: You might want to save taking RDF out for next release
... Don't we have an issue on this
... Have we moved to bugzilla?

MC: We are in transition

JS: We are going to get some bugzilla to IRC support

RS: We talked about it at FTF meeting

MC: Creating an action item
... Section 5 needs major edits

RS: Also appednix

MC: Plan was to use the side by side until tracker issues are closed
... Other groups need to look at both
... We won't worry about RDF for publication, just your comments
... Current editors will be credited as a former editor

JS: Editor emartitus

RS: There are other people who should be included
... need to be pointing to SVG 2, where aria-describedat will be

MC: I am going to redo status
... will add a new status and check with at least RS and JC

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

Clown: This is what I am looking at it
... How user agents should support keyboard...

RS: I am finding stuff we need to delete

MC: Mapping guide does need an edit

RS: There is a lot of stuff that is not interesting anymore
... The whole introductions need to be synchornoized
... I should put the SVG thing in github, but it is ugly

Clown: Send me an e-mail with what you want changed

RS: What time is your meeting tomorrow?

Clown: 3:00pm eastern

RS: Michael do have enough to get this published?

MC: I have to go over some things in the mapping guide, will not get done this week
... The title "Core accessibility API mapping"?
... That is not in the document yet

Clown: i will change it
... Change to "Core....

MC: Do we want the 1.1? Started drafting ....
... I think we still want to have 1.1

Clown: Fine with me

JS: Correct

MC: I have a whole para in the status....
... I will need a summary of what is new
... If you have the changes that will be helpful
... What do we want to ask the public about?
... I always what to help people focus their reviews

<clown> MichaelC: https://github.com/w3c/aria/commits/master/implementation ?

RS: There is not a huge about a stuff we want comments on
... What about role=none?
... Alot of role=presentation out there
... TABINDEX on tabs
... ANy other thoughts?

<richardschwerdtfeger> http://lists.w3.org/Archives/Public/public-pfwg/2014Jun/0000.html

Clown: the GNOME a11y team want updated API mappings, we can ask if I got it right

RS: Editorial, number 4, why gird shows up as "red", some other elements are red

Clown: If you mark them as code they come out red
... it might be the code tag or class name

RS: Section 5.4

MC: Code element
... We can over write the style...

more styling discussion....

MC: Might take a careful code review
... The short name shows up in the URL

RS: Is it all caps

MC: I forgot what we decided

RS: What is that section called...

MC: Reference within the document and short name for the W3C, not rules, my pref is lowercase, but we can do upper case

RS: You mean in the short name

<clown> +1 for lower case in W3C url.

RS: I am with Joseph

<richardschwerdtfeger> +1 for URL’s as lower case

MC: I will make a note, to use upper case for biblo..
... For the mapping specification, we need to send a transition request, we need to get it on record, it might be hard to do this week, next thursday is more likely

RS: its fine with me for next thursday

MC: I am wokring in a branch right now

RS: Any objections for publishing next thursday?
... Do we need to discuss next week?

MC: I will be at the advisory meeting next week, so I don't have anything specific

RS: What changes will be done by wednesday

MC: title, status section, short name....
... I am working on a branch, you can also work on it
... Thats it for the spec
... Ratify on wednesday

JS: Confirm on wednesday for pub on 12th

MC: Can we get the confirmation now?
... Want to to confirm by 5th

<richardschwerdtfeger> +1

RESOLUTION: Request to transition Core Accessibility API Mappings 1.1 to FPWD and WAI-ARIA 1.1 for updated Working Draft

<bgaraventa1979> yes!

<richardschwerdtfeger> +1

RS: Next

<clown> +1

Issue 587

<clown> issue-587?

<trackbot> issue-587 -- Consider allowing the aria-selected state on any focusable element, or add a new attr like aria-active or aria-current -- open

<trackbot> https://www.w3.org/WAI/PF/Group/track/issues/587

RS: We are looking at aria-current and not use aria-select
... ATs don't want to hunt for these
... The discussion is how to move forward
... This will take an ID, and ..... shadow DOM
... aria-current is on a container...

CS: question

RS: you could have an aria-owns and it could be outisde the container

CS: Something other than explicit relationship

RS: It would be on a container and then there would need to be a reverse
... this seems like a new relationship

<joanie> jongund: aha. That's Cynthia talking; not me

Cynthia: I want something other that an ID ref, like ...

CS: If you have a progress diagram, the container is a progress, ...

RS: That was a step within a progress, the role=step is a descendant of role=progress, it is in a progress container

Joanie: I am not sure I understand it, it is preferable than hunting

RS: Given this type of role there can only be one type of container
... A good example is listbox, it can only have options
... Option can only be inside of a listbox, so the container is known
... Is that enough

Joanie: I am not trying to be difficult, ...., steps along a process, it may be hard to find where it says

RS: We don't have one of those left, like "breadcrumb"

<bgaraventa1979> +q

RS: navigation region is a good example, in that case you would need a relationship
... Is there anywhere we can just use aria-current?
... I have seen navigation with tree widgets....

CS: What is more common, a column with 12 links....

RS: In that case you can go back to the container, is that OK Joanie?

Joanie: What I need to know what is the current for?
... Can we provide an ROLE, go from the current object ....

CS: There can be several roles that can have a current

Joanie: I go up until I find the role

<clown> <div role="step" aria-current="true" aria-context="navigation">?

<clown> where aria-navigation takes a role value.

RS: The definition of aria-current, would it be better to say it only applies to specific roles

Jaonie: as long as it is not on too many roles

BG: There is also pagination......
... Have essentially images are being used as part of authentication
... There are many scenarios that current could help, outside of navigation

RS: The current should look for container elements, do we need currentfor?

CS: I think they are different use cases

RS: I think we will need to expand on the use cases, it will require some API changes in the platforms
... 1. we have an aria-current which applies to an element in a navigation container
... 2. aria-currentfor that would apply to any container, takes a single idref

CS: Makes sense to me

<clown> https://www.w3.org/WAI/PF/Group/track/actions/1442

Clown: Look at Issue-1442
... Look at action-1442

<Zakim> Joseph_Scheuhammer, you wanted to note that Bryan G. wrote spec text for aria-current in ACTION-1442.

RS: should it be called called aria-currentfor?
... What about context....

BG: What if there is nesting

We need to have similar names

Clown: owns is for parent child relationships

BG: If you are referencing a parent element from ...

RS: that is a different use case, new action item?
... mixing two different things in the same action
... Can we change this one to currentfor and then create a new action
... I think we have enough for JC to write spec text,

CS: As long as the action item....

RS: Can I edit the cation BG

BG: yes
... I think JC was on the callw e assigned the action item

RS: I think we changed it to currentfor...
... Let me put this text in heere
... We are at the end of time
... Assign to LW?

LW: I would be happy to give it a shot

RS: I could do it, but I am swamped, I know everyone else is too
... I will see if you are in the list, I will assign to LW

zakim scribe

<scribe> scribe: jongund

<joanie> +1 to having the text first before trying to map it

- DRAFT -

Protocols and Formats Working Group Teleconference
02 Jun 2014

Attendees

Contents

Summary of Action Items