- From: Janina Sajka <janina@rednote.net>
- Date: Mon, 21 Oct 2013 11:43:34 -0400
- To: W3C WAI Protocols & Formats <public-pfwg@w3.org>
Minutes from the ARIA Teleconference of 21 October are provided below as text, and are available as hypertext at:
http://www.w3.org/2013/10/21-pf-minutes.html
W3C
- DRAFT -
Protocols and Formats Working Group Teleconference
21 Oct 2013
Agenda
See also: IRC log
Attendees
Present
Michael_Cooper, Janina_Sajka, Joseph_Scheuhammer, Jon_Gunderson, James_Craig
Regrets
Rich_Schwerdtfeger
Chair
SV_MEETING_CHAIR
Scribe
jongund
Contents
* Topics
1. UAIG issues
2. UAIG Testing
* Summary of Action Items
_________________________________________________________________________________________________________________________________________________________________
<trackbot> Date: 21 October 2013
<clown> zakim I am Joseph_Scheuhammer
<clown> http://lists.w3.org/Archives/Public/public-pfwg/2013Oct/0015.html
<MichaelC> scribe: jongund
MC: User agent stuff first
UAIG issues
JS: We need JC, I want to make sure the changes were sufficient
... I can e-mail him I guess
... I have a question for RS, but he is not here
Janina: RS is gone all week
MC: The latest time line, next week is the finailization next week
JS: It puts pressure on the UAIG group if it has to be done next week
Janina: Stack up the best understanding and seek approval
JS: There is a debate with SF
... RS is telling SF no, the consensus is not to change anything
Janina: The change bar is high
JS: There is another JC issue
<clown> http://www.w3.org/WAI/PF/aria-implementation/#keyboard-focus_tabindex
<clown> When the user triggers an element with a defined activation behavior in a manner other than clicking it, such as by pressing Enter, simulate a click on the
element.
JS: JC this should be at risk, since only Opera implements, but the text is in HTML 5, it is at risk there then too
Janina: They have marked the parts that are at risk
JS: Is that the TR?
... That is 5.0
MC: Activate the element as it is clicked
JS: Steps to simulate the click,... creating a click event simulation
MC: This text is a hold over from DOM activate
... We were saying that if you send an activate, also send a click
... DOM has deprecated activate, so should be no need to simulate a click event, since "click" is the new activate event
<clown> www.w3.org/TR/2013/WD-html51-20130528/dom.html#interactive-content
MC: I think we could remove from the guide, and talk tot he HTML5 group
Janina: There is a color coded view of the at risk view
JS: They go into much greater detail about simulating a click
Janina: I will find you the right draft
MC: Sounds like a formulation of what we just talked about
... Are we sure only Opera implements it
... If you press on return on link it follows it on other browsers
... I am not even sure we do need to remove this
JS: Just run the tests and see what happens
... Key is finding a activation behavior
JC: I did the test on OS X FF, Chrome, Safari, and Opera, and out of those, only Opera on the mac support the return key
JS: I thought I saw it on chrome
<janina> http://www.w3.org/html/wg/tests-cr-exit.html
JS: I know DB I discussed it in FF development version
... The green are accepted as passing
<clown> http://www.w3.org/TR/2013/WD-html51-20130528/dom.html#interactive-content
<clown> he user agent should allow the user to manually trigger elements that have an http://www.w3.org/TR/2013/WD-html51-20130528/dom.html#activation-behavior, for
instance using keyboard or voice input, or through mouse clicks.
<clown> When the user triggers an element with a defined http://www.w3.org/TR/2013/WD-html51-20130528/dom.html#activation-behavior in a manner other than clicking it,
the default action of the interaction event must be to http://www.w3.org/TR/2013/WD-html51-20130528/dom.html#run-synthetic-click-activation-steps on the element.
JC: I think it should be in HTML, THAT IS WHERE IT SHOULD BE
JS: Why is it green
Janina: It means it has been accepted as interoperable
JC: There may have been implementation in the last month
Janina: I saw we close this
<clown> issue-616?
<trackbot> issue-616 -- ISSUE: Review potentially at-risk statement "When the user triggers an element with a defined activation behavior in a manner other than
clicking it, such as by pressing Enter, simulate a click on the element." -- open
<trackbot> https://www.w3.org/WAI/PF/Group/track/issues/616
JC: Say it is an HTML 5 requirement
MC: Then we don't have to test
JC: Try a div with an onclick event handler
... Things that don't have, just have been defined as ARIA controls
JS: Like role=link?
MC: it does not work role=link or role=button
... This is necessary to get ARIA feature to behave as native features
<clown> http://www.w3.org/WAI/PF/aria-implementation/#def_activation_behavior
MC: Do we intend this to apply to all ARIA?
JS: It is not very aluminating
MC: Defined activation behavior ....
... This does mean that an ARIA role on a DIV, either the AT needs to send to click event, or authors need to know how to respond
JC: AT is not the problem here
... the problem is the full keyboard activation
... For ARIA 1.0 authors need to capture those keys
... Otherwise we would be changing the default behavior of the elements
MC: I am convinced by that, we need to reopen the question in ARIA 1.1
... We should remove from 1.0 and put an issue in for ARIA 1.1
JS: CS will say let it stay there
MC: So it does look like we can get test able statements
... It is an easy enough test case, we are pretty sure it will not pass
JC: My concern, even if wedo get implementation, some authors will not expect this behavior
... We could add an add activation behavior
MC: I like removing this from 1.0 and reopening in 1.1
Janina: Seems to be reasonable approach, if we do something in 1.0 will be more a dance
JS: I will try to come to UAIG tomorrow
MC: I will be there tooo
JS: She said this should have done alone time ago
Janina: One person objecting does not break consensus
MC: Groups can proceed over the objects of a single person
... The benefit does not justify the cost, we will have more time in 1.1, we only recently identified this problem
JS: COllectively remember these discussions
MC: Any other UAIG issues?
<clown> action-1262?
<trackbot> action-1262 -- Joseph Scheuhammer to Update 4.1. focus states and events table to clarify differences between platforms and dom/desktop/at focus.
http://www.w3.org/wai/pf/aria-implementation/#focus_state_event_table -- due 2013-09-23 -- PENDINGREVIEW
<trackbot> https://www.w3.org/WAI/PF/Group/track/actions/1262
JS: Action 1262
JC: Sorry for being late, I am still ill
JS: You should be in bed
<clown> http://www.w3.org/WAI/PF/aria-implementation/#focus_state_event_table
JS: I added content to address the focus issues you raised, these have been vetted with CS and DB
JC: Give me a second to read them
... I think it is OK
... Second bullet is a little confusing
... In recent discussions with hixie about DOM focus and activation focus are different..
... need more explanation, like for example .....
... I am not adverse to putting in a note
JS: I am not adverse to putting in a note
JC: I am not sure a note is strictly necessary
JG: Sometimes it is advocates that are reading this document, to help with developers
JS: I can do it
JC: Since it is a note it is editorial
Janina: We need to close stuff this week, it we will meet this pblicatoin date
<clown> action-1269?
<trackbot> action-1269 -- Joseph Scheuhammer to Prefix bullet 2 of #keyboard-focus_aria-activedescendant with "for platforms that expose an accessibility focus
separately from the keyboard focus," -- due 2013-10-07 -- PENDINGREVIEW
<trackbot> https://www.w3.org/WAI/PF/Group/track/actions/1269
JS: Next pending review, action 1269
<clown> http://www.w3.org/WAI/PF/aria-implementation/#keyboard-focus_aria-activedescendant
JS: JS: Bullet 2 of the active descendent section
... Everything else is identical
JC: I missing the different part
... Ok item 2
... It seems right to me
... Thank you
JSL Your welcome, closing 1269
JS: That's all I have
... For UAIG stuff
Janina: That's the one we are trying to get out the door
JC: There a number of 1.1 issues I would like to start going through
Janina: We need to get this second last call out the door
... Want to make sure that there are no supprises
UAIG Testing
MC: I was on mute, i was talking alot
... RS wants to be present for 1.1 discussions
<clown>
https://dvcs.w3.org/hg/pfwg/raw-file/default/ARIA/1.0/tests/test-files/roles-properties-supported/roles-properties-supported-combobox-aria-autocomplete-none.html
MC: Let's move on to UAIG testing
JS: This is problem with FF, a combo box with aria-autocompletion=none, but when I tested on FF it HTML5 text elements autocomplete is always true
JC: in addition we have a host language feature that conflicts
JS: We have an example that does not work anymore
... Input type text in HTML 5 always has autocomplete is true
JC: Most browsers render the content independent of the doc type
JS: I tested on ATK-SPI
MC: this is one of the test files that are shared
... We have our passes, we shoudl add it to the ARIA 1.1 issue list about the auto complete conflict with aria attribute
JS: You are going to get auto complete on spinners using text boxes
<jcraig> ACTION: jcraig to note in ARIA 1.1 spec issue to note on aria-activedescendant that HTML5 has a a native autocomplete attr that may be in conflict with this
one. [recorded in http://www.w3.org/2013/10/21-pf-minutes.html#action01]
<trackbot> Created ACTION-1281 - Note in aria 1.1 spec issue to note on aria-activedescendant that html5 has a a native autocomplete attr that may be in conflict with
this one. [on James Craig - due 2013-10-28].
JS: The cure is to set auto complete equal to false it they do not want autocomplete behavior
<MichaelC> https://www.w3.org/WAI/PF/testharness/testresults?testsuite_id=1&testcase_id=172
<jcraig> ACTION: jcraig to file HTML-AAPI bug related to mapping native @autocomplete to @aria-autocomplete in strict vs weak semantics tables. [recorded in
http://www.w3.org/2013/10/21-pf-minutes.html#action02]
<trackbot> Created ACTION-1282 - File html-aapi bug related to mapping native @autocomplete to @aria-autocomplete in strict vs weak semantics tables. [on James Craig
- due 2013-10-28].
Janina: If they passed we are done, even if the implementation has changed
JC: A 1.1 issues are you talking about UAIG?
MC: I am not sure, but probably starts in the 1.1 spec
... We may want to bring up witht he HTML working group
JC: i am going to file a bug, there is a table identifies strong and weak native semantics
MC: make default that auto complete is by default false
Janina: The document is a shared document, so we are required to approve changes
JC: It works better if we don't make a big deal and to work with SF as more an editorial issue
Janina: Way to start
... We are part owners
MC: We need to sign off before publication
JS: This includes heart beat publications
... There are a bunch of tests for UAIG? There was suppose to be one for ATK-SPI, i could not find
MC: I don't see it now
JS: OK
... That's why I did not find it
MC: We are removing the activate simulate click requirement, removing the tests
... We have less test cases
JS: i don't know what DB and CS are doing, I will bring it up tomorrow
MC: We need someone who can test
JS: We can ask them
... We need to create 7 test cases and test them on multiple OS in the next week
Janina: What do we want to try to achieve tomorrow?
JS: David said can he do this through e-mail or other channels
... has anyone heard from him?
Janina: Do it on the call?
JS: We did that last week
... CS thought her actions were done
Janina: Expected results would not get us there
JS: they should do it on a call
... Go through the wiki page and clarify what they mean and write the test case
MC: I thik we should try to make that happen tomorrow
... If we can get it done in one call, then we do not need to bug them
JS: How many for OS X?
MC: There is one specific for OS X
... Live region tests, with a table...
JC: What is the test case number?
MC: There is no test case yet
<MichaelC> http://www.w3.org/WAI/PF/aria-implementation/#mapping_events_visibility
MC: There are requirements in a table for the behavior
JC: That is under 5.8.2
<MichaelC> http://www.w3.org/WAI/PF/wiki/ARIA/Tests/Testable_Statements#mapping_events_visibility
MC: the closest header is here
... A test case and file if it is needed
JC: I don't see the @@
MC: the second link
<MichaelC> tests 72 to 84
JC: If you should a DOM node, and change the CSS to display: block
... So it is not necessarily part of ARIA 1.0
MC: The tests are 72-84,
Janina: Does that take our list from 7 to 6
MC: It might have
... I need to check if the UL is the right place for events, i am not sure what that means
<jcraig> ACTION: jcraig to add ARIA spec live region test cases 72-84 to the ARIA Implementation Guide tests, and run those test cases for Mac [recorded in
http://www.w3.org/2013/10/21-pf-minutes.html#action03]
<trackbot> Created ACTION-1283 - Add aria spec live region test cases 72-84 to the aria implementation guide tests, and run those test cases for mac [on James Craig -
due 2013-10-28].
JS: Table talks about the accessibility sub tree
MC: If we don't have any problems we can probably delete
JS: So you have added a new action
MC: I think the question there, those tests are working, I think we can remove that, so i am doting that now
... One of the @@ is on....
... I am going to remove that @@
JS: The whole working in that table has changed, it said like "these events may be trimmed", if you have multiple selections, user agent may trim out some of the
events for performance reasons
MC: I wil post, but not sure if the test results are good
JS: that entire row has nothing but "may" in it, there are no "musts"
Janina: Another one we do not have to worry about
JC: Can someone read the requirement, i am outside
JS: The implication is that the user agent is ....
Janina: What is the verb for apple
JS: It is an implied must
Janina: We need some TLC from JC on that one
MC: the test case is written
... There will be a test case in a little while
JC: I can write it
MC: i will try finishing it, but if I have a problem I willsend t to you
... there are six @@ now
JC: I am going to have to leave soon
Janina: Get well
MC: 5.1.2 including elements in the accessibility tree
... reviewing spec requirements, looking at coverage from other specs and test suites
JS: there are only 6
MC: I think we need to take a closer look at this section
<janina> rrsagent make minutes
Summary of Action Items
[NEW] ACTION: jcraig to add ARIA spec live region test cases 72-84 to the ARIA Implementation Guide tests, and run those test cases for Mac [recorded in
http://www.w3.org/2013/10/21-pf-minutes.html#action03]
[NEW] ACTION: jcraig to file HTML-AAPI bug related to mapping native @autocomplete to @aria-autocomplete in strict vs weak semantics tables. [recorded in
http://www.w3.org/2013/10/21-pf-minutes.html#action02]
[NEW] ACTION: jcraig to note in ARIA 1.1 spec issue to note on aria-activedescendant that HTML5 has a a native autocomplete attr that may be in conflict with this
one. [recorded in http://www.w3.org/2013/10/21-pf-minutes.html#action01]
[End of minutes]
_________________________________________________________________________________________________________________________________________________________________
Found Scribe: jongund
Present: Michael_Cooper Janina_Sajka Joseph_Scheuhammer Jon_Gunderson James_Craig
Regrets: Rich_Schwerdtfeger
People with action items: jcraig
--
Janina Sajka, Phone: +1.443.300.2200
sip:janina@asterisk.rednote.net
Email: janina@rednote.net
Linux Foundation Fellow
Executive Chair, Accessibility Workgroup: http://a11y.org
The World Wide Web Consortium (W3C), Web Accessibility Initiative (WAI)
Chair, Protocols & Formats http://www.w3.org/wai/pf
Indie UI http://www.w3.org/WAI/IndieUI/
Received on Monday, 21 October 2013 15:43:58 UTC