See also: IRC log
<joanie> agenda
<joanie> s/:this//
<ShaneM> yes please
<scribe> scribe: JF
<clown> https://www.w3.org/wiki/ARIA_1.1_Testable_Statements
stefan: edited table, work is progressing
RS: when do you think the tests will be completed?
Stefan: hopefully next week
JD: there are still about 130 left to do
RS: we need more help - anyone
else able to assist?
... we are running late: we need these test examples done. We
need more help
... asking if BG can take on some test.
BG: not clear on how to deliver. Is there a how-to guide?
<clown> https://www.w3.org/wiki/ARIA_1.1_Testable_Statements
RS: for every testable state, we are creating an example. Very basic, doesn't need CSS (etc.) - just expected results
<clown> https://www.w3.org/wiki/ARIA_1.1_Testable_Statements#Introduction
then there is a table for each platform, for a given element
that needs to be able to prove the assertion
we have an element with an id that is affected, and then for each platform what the mappiings are
[rich shows an example]
<richardschwerdtfeger> if given
<richardschwerdtfeger> <div role="button">
<richardschwerdtfeger> OK
<richardschwerdtfeger> </div>
<richardschwerdtfeger> then for the element of id test expose the button role but do not expose aria-haspopup
find associated element, then get the platform APIs and then report results
for any of the testable statements, we should be able to determine success or failure fairlyeasily
[Rich assigning some tests to Brian G.]
JD: wonders if we should also look at aria-separator and ...
<joanie> https://www.w3.org/wiki/ARIA_1.1_Testable_Statements#aria-expanded
JD: link to section of testable statements - aria-expanded
there are 3 test for seperator
there is a question regarding separator
the defualt for aria-expanded is undefined, thus not mapped
as a dev, this is complex - I don't want the mappings to have too much info
what it says in the core AAM and what actual tech is delivering, is what we need to test for
<joanie> https://lists.w3.org/Archives/Public/public-aria/2017Mar/0017.html
Rich feels however that user-agents should do it
RS: dealing with W3C site issues
- won't allow to assign Brian G. his tasks
... I posted something to the list]
JD: see the link I posted - it is Rich's post (which we are discussing now)
RS: if we start down the path where somebody applies inaapropriate attribute to a role
and then we ask the browsers to not map it... that will be very expensive to the browsers
as they will then need to ask if it is accurate
and some may not be able to do it
if you apply an attribute to a given element thta is incorrect, many tools flag the errors
RS: Joanie is arguing that if we apply this... if we don't support separator expanded
Joanie wants to check to see if it is correct or not, and then we catch it
RS: but that then means that the user-agents will need to test for that
understand that it was previously supported, and that may make this an edge case
but if we start down this road, there is a concern that we'll get push-back from the browsers
there is a potential that this will set us back (re: CR and our larger timeline)
if we go to CR, we need a high-level confidence that browser vendors will implement all of this. This could cause the core AAM from being sent back -aria is dependant onthat
JD: don't we have to test for all normative changes?
RS: yes. but we are talkng about a large change - we will be asking browser vendors to do a lot
we may get away with this once, but...
RS: we seem to be asking the browsers to become both test and repair tools
JD: largest concern is the CR status - if a change is normative, we have to test it (point finale)
RS: my suggestion is to move the tests to the core AAM section of the test harness
and find out quickly (and avoid scope creep !!!!)
if Joanie can get Alex @ Moz to make the change - then great
but if Moz isn't prepared to do so, that we yank it
all it says is that it is not supported
Clown: if somebody puts in something that is "illegal" it will likely be exposed int he Accessibility tree, but that's it
RS: I believe that to be true as well
Clown: but Joanie still has a point
RS: agreed, but it may set a bad precedent, may create major push-bacjk
however, if we've removed it - we should put the suppressionpart under the core AAM. The author tools and the testing tools shold get it
JD: will need to review more code
however, I found an example (in Webkit) methods that say "is this really supported" on a specific role. Not sure about Gecko however
RS: Gecko allows a lot of stuff to be put on a lot of things
JD: Orca has a ton of stuff that "ignores" things from Gecko
RS: so, group think?
JF: I hear, pursue this as far as possible, but be prepared to drop it if it impacts the ARIA 1.1 timeline
JD: I will live with whatever the group wants
RS: so, do we put this into the core AAM testable statements... Joanie will track this down fairly quickly
any objections to parking this at AAM Core?
<jongund> OK with me
hearing general consensus
RS: so it's still in until it's
out :)
... passes around the volunteers search (again)
[discussion around getting @clown more time to help here]
RS: will try and gather up some more help off-line
JG: made a lot of progress at
CSUN - some good meetings with info for other platforms
APIs
... have 2 students helping who seem to be able to run the test
platform
starting a process of adding iAccessible 2 support
also trying to take joaine's code and modifying that to leaverage it
hoping to connect with Joanie and students for a "tour"
JD: sure, but it may be more complex than just replacing one thing with another
JG: don't think it will be that easy - realistic expectations, but want to better understand what is happening
JD: a lot has changed
JG: OK, well...
[Joanie and JG talking about testing needs and emergent work]
JD: also chatted with Shane M
last week, and he offered some suggestions
... no matter what, don't think her code will be that re-usable
by Jon
JG: still would like to do a
deeper dive and have those chats
... hopeful to get stuff built sooner than later, but was
hoping to leaverage more of Joanie's code than it now
appears
JD: have been actively working on her core platforms APIs - hope to show progress next week
zzakim, take up next item
JD: MSFT has volunteered tester help, but no news on ATTA... may need volunteers for windows testing
JG: there may be students that ccould help with taht - may not be in a good timeframe however
JD: I think I may be able to reuse a lot of my code... starting to already fill in the blank
*HOPE* to be able to do it, but cannot promise
JG: we have people who can do
manual testing (on Edge)
... we've done that in the past
RS: that may be OK for UIA
<richardschwerdtfeger> https://www.paciellogroup.com/resources/aviewer/
accprobe has not been updated for IA2
S. FAulkner looking at updating aViewer
<richardschwerdtfeger> http://pfiddlesoft.com/uibrowser/index.html
there is another tool for the mac platform too
<richardschwerdtfeger> for mac
Clown: for safari? what does it mean "for mac"?
RS: not sure -
investigating
... cannot do any manual testing until computer is re-installed
locally
JD: will aViewer be the testing tool for UIA?
RS: not sure yet
<joanie> The Accessibility Viewer (aViewer) is an inspection tool for Windows that displays the accessibility API information (MSAA, IAccessible2, UI Automation, ARIA, HTML DOM) exposed by web browsers to the operating system, and thus to any assistive technology (AT) such as screenreaders.
JD: this sorta looks like... if Jon has volunteers who could do UIA testing could start using this tool
<joanie> https://github.com/ThePacielloGroup/aviewer
[discussion if it is a tool that is open source]
JG: thinks it is a wrap-around some Java thing
looked at this previously, and it is a compiled binary - there is no code to build upon
JG: if I knew more about IA2, I'd be able to better forecast a timeline
best i can say is taht I have a lot of folks lining up to offer help, so full speed ahead into the unknown
RS: if aViewer is updated and
adequate, do we start manual testing ASAP?
... swome W3 pressure on "when will we be done?"
JG: hopefully will be better informed by next week. Will lean on Joaniefor some direction and assistance
[quick discussion around SVG accessibility]
RS: there are concerns over the SVG WG at W3C... need to follow up there
trackbot, end meeting
This is scribe.perl Revision: 1.152 of Date: 2017/02/06 11:04:15 Check for newer version at http://dev.w3.org/cvsweb/~checkout~/2002/scribe/ Guessing input format: Irssi_ISO8601_Log_Text_Format (score 1.00) Succeeded: s/clear agenda// Succeeded: s/agenda// Succeeded: s/agenda: be done//g FAILED: s/agenda:this// Succeeded: s/?// Succeeded: s/agenda: this// Succeeded: s/agenda// Succeeded: s/JB/JD/ Succeeded: s/clown: edited table, work is progressing/stefan: edited table, work is progressing/ Succeeded: s/failryl /fairly/ Succeeded: s/puch-back/push-back/ Succeeded: s/JB/JD/g Succeeded: s/ORCA/Orca/ Succeeded: s/ATIA/ATTA/ Succeeded: s/jaonie /Joanie/ Default Present: Joanmarie_Diggs, Stefan, MichielBijl, janina, Joseph_Scheuhammer, jongund, Joseph_Scheuhamer Present: Joanmarie_Diggs Stefan MichielBijl janina Joseph_Scheuhammer jongund Regrets: James_Nurthen Matt_King Michael_Cooper ShaneM Found Scribe: JF Inferring ScribeNick: JF Found Date: 09 Mar 2017 Guessing minutes URL: http://www.w3.org/2017/03/09-aria-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]