Protocols and Formats Working Group Teleconference -- 23 Sep 2013

<trackbot> Date: 23 September 2013

<richardschwerdtfeger> meeting: W3C WAI-PF ARIA Caucus

<janina> trackbot, start meeting

<trackbot> Meeting: Protocols and Formats Working Group Teleconference

<trackbot> Date: 23 September 2013

<richardschwerdtfeger> calling back in

<janina> SIP is definitely not working for Zakim today.

RS: Do you know what changes there have been to the document

<richardschwerdtfeger> http://www.w3.org/WAI/PF/aria-1.1/

MC: I can't find the changes right now
... The list of changes is not very informative

<MichaelC> http://www.w3.org/WAI/PF/aria-1.1/appendices

MC: I will regenerate

<janina> scribe: jongund

<clown> http://www.w3.org/WAI/PF/aria-1.1/appendices#changelog

MC: there are just 2 normative changes since t =he last public working draft
... I will fix that late today

RS: Added none default false to aria checked state

JS: That is interesting working, that is the default
... What this now says, if you look at the checkbox role, implicit value .....

RS: I am looking in the aria spec, the element does not support being checked

JS: But that does not make sense for checkbox

<MichaelC> Diff between ARIA 1.0 and ARIA 1.1

RS: He changed the value, changed the defualt ...

<MichaelC> TOC issue fixed, changes logged at http://www.w3.org/WAI/PF/aria-1.1/appendices#changelog

JS: This goes back to the April FTF discussion
... If you leave it off, the checkbox checked will be false
... Combobox has a required expanded state and that will be false

JG: I don't like it, since validators don't know if the developer
... I will never like
... It is just one more thing to live with

RS: Any objections?
... We will keep the resolution for missing aria-checked

<clown> http://www.w3.org/WAI/PF/aria-1.1/roles#combobox

RS: Are we missing something for combobox?

JS: it is in there
... Implicit value for role....

RS: The implicit value for....

JS: Combobox role, very lat role, implicit value for role
... Added default for aria-expanded to be false

RS: I am OK with that
... Do you expect UAIG to expose..

JS: They all ready do, and we have test cases

RS: Do we put a collapse state out there?

JS: It will expose as expanadable and collapsed

RS: James?

MC: I was rejoining

RS: Who has the pen to not the change?

MC: I have the pen, what do i need to change
... I don't see anything in the diff version

JS: Does that mean it is in ARIA 1.0?

<clown> http://www.w3.org/WAI/PF/aria/roles#combobox

JS: It is not in the published draft, maybe in the editors

MC: He made ARIA checked ....
... Let me go back to my diffs

<clown> http://www.w3.org/WAI/PF/aria/roles#combobox

<clown> http://www.w3.org/WAI/PF/aria-1.1/roles#combobox

JS: There is 1.0 and 1.1

MC: That did not get documented, he did on 22 spetember

JS: I will get the source up

RS: We will add the following text....
... Should that have a number next to it

MC: I added this

<richardschwerdtfeger> Add the following text to section 10.3:

JC: What are the differences between 1.0 and 1.1, the biggest difference is aria-describedat
... We have not done a rework of 1.0, so it should not be a difference 1.0 and 1.1

MC: The differences are in the editors source, but not in the public draft
... I will update the make file after testing , later this week

<richardschwerdtfeger> http://www.w3.org/WAI/PF/aria-1.1/states_and_properties#aria-describedat

RS: We went though the default values
... working on aria-describedat now
... SHould we make that a note?

JC: A note or something else, it is something we should definitely add

MC: Maybe we should be referencing .....

JS: We would like to says yes to MC thought

RS: Talk to URI people

MC: the do to ..., looking for URI specification..

<richardschwerdtfeger> In order to provide input- and device-independent access to the associated descriptive content, authors SHOULD use the tabindex attribute to ensure the object is included in the default tab order.

RS: input dash, what is input dash
... Just trying to clarify

JC: that is how you share ... input and device independent ..., grammatical short hand

RS: That will have to change in the future

JC: Why do we have to change that

RS: They will be doing it in SVG to

JC: That will not affect us

RS: In a later version of SVG you may want to change it

JC: if you take out XLINK it doesn't work
... If SVG2 will still need to be compatible with current versions
... We can say this is an SVG 1 document

RS: that's fine
... I am ok with this

MC: I have been working on a refinition of URI that is less HTML centric

<MichaelC> proposed for URI: A Uniform Resource Identifier as defined by <a href="http://www.ietf.org/rfc/rfc3986.txt">RFC 3986</a> [RFC3986]. It may reference a separate document, or a content fragment identifier in a separate document, or a content fragment identifier within the same document.

MC: I think we should put in value type
... Propse this for the URI, I am not sure where to put processing instructions, but in general should be part of the host language

JC: Is that a to do?

MC: Took that out

<richardschwerdtfeger> Example: Using aria-describedat to provide a detailed description for raster images displayed in SVG

RSL You have an SVG underline, did you want to link to the SVG document

RS: Do we have a glossary

<richardschwerdtfeger> [SVG]

<richardschwerdtfeger> http://www.w3.org/WAI/PF/aria-1.1/references

JJC: that is a good point

RS: We have an SVG in there, reference that and you will be good
... Anything else for the good of the order?

MC: Actually I need to get a working group consensus

JC: I have a question
... Can we add the text role or disclosure role....

MC: I don't think we can by this thursday

JC: Ok this thursday

MC: We can do a second public working draft any time

JC: Will you make the changes?

MC: Yes

JC: I will give you the actions

<richardschwerdtfeger> http://www.w3.org/WAI/PF/aria-1.1/appendices#ack_others

MC: I am planning to do a pass

<richardschwerdtfeger> http://www.w3.org/WAI/PF/aria/appendices#acknowledgements

<jcraig> ACTION: cooper to finish describedat: add [svg] ref, change todo to note or normative text re: "authors and book publishers should use native markup..." [recorded in http://www.w3.org/2013/09/23-pf-minutes.html#action01]

<trackbot> Created ACTION-1265 - Finish describedat: add [svg] ref, change todo to note or normative text re: "authors and book publishers should use native markup..." [on Michael Cooper - due 2013-09-30].

<richardschwerdtfeger> http://www.w3.org/WAI/PF/aria-1.1/appendices#ack_others

RS: Can we make it a list

MC: I understand, there are no standard, the primary list and a also contributed list

JS: Why is it numbered, rather than numbered
... It looks like it is alphabetical

RS: Anyone we need to add to this list?

JC: Alex Chin and Alex Surkov ...

RS: Let's see if we missed anyone, what about George

MC: He is not a member, but he is in the second list

<richardschwerdtfeger> George Kerscher

<jcraig> ACTION: Cooper to move Alex Surkov and Scott Gonzalez from Other list to Participants list [recorded in http://www.w3.org/2013/09/23-pf-minutes.html#action02]

<trackbot> Created ACTION-1266 - Move alex surkov and scott gonzalez from other list to participants list [on Michael Cooper - due 2013-09-30].

<clown> Scott González (JQuery Foundation),

MC: What is his affiliation?
... Daisy

JS: IDPF

<jcraig> action-1266?

<trackbot> action-1266 -- Michael Cooper to Move alex surkov and scott gonzalez from other list to participants list -- due 2013-09-30 -- OPEN

<trackbot> https://www.w3.org/WAI/PF/Group/track/actions/1266

RS: Is steve faulkner a member of PF
... Just want to make sure we give people credit

MC: Steve is in the list

RS: Is he still an invited expert?

MC: yes

RS: Did John Foliot ever join?

MC: no, but not for lack of effort
... I am convered this to a unordered list

RS: Anything we need to do with the status of this document

MC: I cannot post a version with a status, but I can tell you

<MichaelC> Proposed Status:

<MichaelC> This is a First Public Working Draft of WAI-ARIA 1.1 by the Protocols & Formats Working Group of the Web Accessibility Initiative. The primary feature that is new since WAI-ARIA 1.0 is the aria-describedat property, which aims to meet extended description use cases that are not met by aria-describedby. The specific engineering of this feature may change and is being published now to collect public feedback both on this proposal and suggestions for other proposals

<MichaelC> . Future drafts of ARIA 1.1 will also provide a small number of features to complete the HTML + ARIA accessibility model. It is expected this will complement HTML 5.1. ARIA 1.1 will have a very limited scope. Requirements will be published in the near future. This version of ARIA 1.1 also includes a small set of editorial changes that accumulated in ARIA 1.0 after the publication of the Candidate Recommendation. A history of changes to WAI-ARIA is available.

<MichaelC> For this publicaiton, the Protocols and Formats Working Group primarily seeks feedback on the aria-describedat feature, but feedback on any aspect of the specification is accepted. To comment, send email to public-pfwg-comments@w3.org (comment archive). Comments are requested by 1 November 2013. In-progress updates to the document may be viewed in the publicly visible editors' draft.

MC: There some boiler plate
... the thrid chunk is the start of a new paragraph

<clown> +1 to the proposed status as stated by MichaelC.

JG: can you change "UIUC" to "University of Illinois"

JC: It is a note about what we have to do

MC: it is more than I can do on the call

JC: I added it to the tracker

RS: Any one object ot promte to a public working draft

RESOLUTION: request publication of ARIA 1.1 as a First Public Working Draft

RS: On to the next set of stuff, JC thank you for getting this done

JC: You're welcome. Thank you.

RS: Next thing is the UAIG issues?
... Anything JS?

<clown> action-1198?

<trackbot> action-1198 -- Joseph Scheuhammer to Update non-default values for required attrs per role. e.g. checkbox and other checkable role default to aria-checked="false" rather than the attr default: undefined. -- due 2013-05-01 -- PENDINGREVIEW

<trackbot> https://www.w3.org/WAI/PF/Group/track/actions/1198

JS: There a couple of things
... Action 1198
... Work that JC did to add aria-expanded default value

JC: that should be done

JS: I understand now that the draft has not been published
... I am closing it then

<clown> action-1219?

<trackbot> action-1219 -- Joseph Scheuhammer to UAIG: passive RFC requirement: two instances of "The state change events MAY be trimmed out for performance" Who is this requirement for? Authors? UA? AT? Be specific. What does "trimmed out" mean? Be specific. -- due 2013-05-14 -- OPEN

<trackbot> https://www.w3.org/WAI/PF/Group/track/actions/1219

JS: this is one last week,and discussed with Alex
... RS you raised this issue, do we need to look at UAIG?

RS: Did you put some text in?

JS: No, Alex said there was nothing to change

JC: Tis may is for user agents or authors

JS: Everything in UAIG is for the user agent
... I could put in the user agent ....
... Could I just put in a para about all requirements are for user agent, might already be there
... You select shift click...

RS: How does user agent detect...

JS: I don't know how i is done, but FF puts out one event if more than 5 individual events

RS: is there a normative wat ....
... If a select 3 will alex fire?
... the state change event needs to be trimmed out ...

JS: Let me get the section up, 5.8.3

RS: Sorry to be a pain, but I am just saying, that the user agent may fire this ...
... State change events maybe trimmed out for performance

JS: I took all of these .....

RS: I don't view these as state change....

JS: Write a email to Alex ....

RS: I understand what he is doing, I am not sure what IE does

<richardschwerdtfeger> The user agent MAY fire a EVENT_OBJECT_SELECTIONWITHIN event. If this happens the state change events MAY be trimmed out for performance.

JS: So you just want to add an if

RS: What is many? It is not consistent

JS: That's why it is a may, if we want a number we can get agreement and make it a must
... What is event object selection?

RS: Selection within is not a stat on the container

JS: Remove ...

JC: Maybe this is the state flux, rather than ...
... It is a binary state

RS: You can have a selected state on an container, how many changes does it take to fire the event, to me is semantics

JS: What is being trimmed out ......
... What is event object ... os being trimmed out too

RS: it is applied to individual events ....

<richardschwerdtfeger> User agents MUST fire the following events when aria-selected changes on a descendant, as follows:

JS: Just ays the UAIG text ....

RS: There a lot of events in here they do not need to be state changes

JS: they say state change event

RS: Something in here and they know when the trigger it

<clown> The other events noted above MAY be trimmed out for performance.

JS: You won't be firing ..... in this column

RS: Other events in this column will be trimmed for performance

JS: That is what I just put in the chat room
... that is the only one that gets fire

<clown> EVENT_OBJECT_SELECTIONWITHIN is all that is necessary. The other events noted above MAY be trimmed out for performance.

RS: The left column says that .....
... That makes sense to me

<clown> User agent MAY fire a EVENT_OBJECT_SELECTIONWITHIN. The other events noted above MAY be trimmed out for performance.

<richardschwerdtfeger> User agent MAY fire a EVENT_OBJECT_SELECTIONWITHIN. If this event is fired the other events noted above MAY be trimmed out for performance.

RS: The Mac column is really easy
... It looks like the only event is object selection change

JS: In the other case you fire many more of them, under certain conditions

RS: on the container?

JS: It does not say that now

<richardschwerdtfeger> The user agent may fire a single object::selection_changed on the container. In this case, the user agent MAY trim out object::selection_changed generated by owned children for performance reasons.

RS: Owned children the correct terminology, owned descendants

JS: It fires on the container

RS: Always on the container?

JS: That's what it looks like

<richardschwerdtfeger> The user agent may fire a single object::selection_changed on the container vs. multiple.

RS: If you want to say for performance reasons

JS: That is the reason
... The last one is about ....

RS: I am not sure it the UIA stuff has been tested

JS: It is not in IE
... The UIA columns are at risk and maybe removed
... I will pass this by Alex
... There are other outstanding issues, but waiting for responses

RS: one thing left

JS: i will try to write some test cases this week

RS: You get a focus and select state change

JS: I gave the new proposed test file and alex said that is the right test
... Let me find the e-mail
... I replied to all, was it on the public PF list?
... What is the public-pfw

<clown> http://lists.w3.org/Archives/Public/public-pfwg/2013Sep/0048.html

JS: it was on fiiday, found it

RS: Did you change it?

JS: Alex was the only one who replied
... I said the test file need to change

RS: Not author driven

JS: Then your not testing an ARIA listbox, you are only testing the select

RS: if one of the items changes the focus then event....

JS: Author must configure ...
... That is one of the changes I made

<clown> Using 85 as the example, assuming it is a test of a multiselectable container (not sure):

<richardschwerdtfeger> :-)

RS: What is the testable statement?

MC: I did not write the testable statement, I just copied
... I am not sure who worte the test case
... This was a single selection use case
... My source

RS: I have another engagement...

<MichaelC> these came from http://www.w3.org/WAI/PF/wiki/ARIA/Tests/Testable_Statements#mapping_events_selection

MC: They were single selection list

<clown> div element with role="listbox" and it is a single selection list, and child div elements with role="option" and tabindex="0", and focus is moved among the options via TAB and SHIFT-TAB keystrokes.

RS: I can put that in here

JS: This is the correct testable statement

RS: We just need to test one

JS: Sure i guess

MC: I think that you need at least 2 focusable, to be able to move among them

JS: If you have selection following focus, then these events apply

RS: The user agent actually exposes ...., that is what FF does

JS: Why was it failing?

RS: I got to run
... Try to hook up this week

JS: i thought this was resolved Friday, but it is not

RS: MC can you chair the next 2 weeks?

MC: Yes

RS: We got our big item done to day

- DRAFT -

Protocols and Formats Working Group Teleconference
23 Sep 2013

Attendees

Contents

Summary of Action Items