Protocols and Formats Working Group Teleconference

17 Dec 2015

See also: IRC log


George_Kerscher, JemmaJaEunKU, LiamQuin, MichaelC, MichielBijl, Rich, ShaneM, TzviyaSiegman, fesch, janina
BryanGaraventa, JonGunderson, LeonieWatson, JoanmarieDiggs
MichielBijl, janina


<richardschwerdtfeger> Meeting: W3C ARIA Working Group

<richardschwerdtfeger> https://lists.w3.org/Archives/Public/public-pfwg/2015Dec/0134.html

<MichielBijl> scribe: MichielBijl

<jemma> can someone copy webex meeting link here?


<jemma> thanks Michiel

Details/Summary and the creation of a role to identify Extended Descriptions

<richardschwerdtfeger> https://lists.w3.org/Archives/Public/public-pfwg/2015Dec/0134.html

RS: problem with link
... it can't be a role, it has to be an attribute
... you still want to know it's a link
... if it's detail/summary you want to know it is a detail/summary

JS: things could be external

CS: is this a whole new thing?
... longdesc isn't stringyfied
... how do I link to external sources on the API level?

RS: longdesc applies directly to an element
... What if the link is elsewhere in the document?
... What if we use details/summary?
... You could reference the details element
... you have an object that reveals the content
... I don't know if it'll work the same as longdesc
... I'm guessing AT can go in and activate the longdesc?

CS: This is not final, but I've been thinking of creating a hyperlink pattern.
... So the element could be whatever it is, but act like a hyperlink

JS: that could still be internal

CS: It's going to be hard enough to implement one mechanism
... let alone two

RS: could it be just describedat?
... You don't want to stringify

JS: There is nothing in the HTML spec that limits it to description

CS: Don't think we need something special

GK: Information on another server can be updated.

<Zakim> clown, you wanted to ask if the role could be applied to what the link points to?

GK: We have description, simplified description, 3d printer item, each has human readable description. And a way to select it.

TS: I have alt text, and a video description, are we talking about described by?
... We can't just invent a new role

CS: I got two places I can put something, so there can be only two things.

<Zakim> tzviya, you wanted to ask if the plan is to use describedby w changes

RS: you can have a lot of stuff inside of details if you so choose.
... We need a navigation mechanism to get to the description.
... We need to say that describedby will stringify the content.
... We don't want it to be stringified whatever the mechanism
... You can put a description in a string, or a relationship.
... We need to tell the UA not to stringify it

JS: What is the common use case? Remote or current document?

CS: Use longdesc

JS: That is only for images

NOTICE: 5 minutes left for this topic

JS: Mozilla loads the contents of the link?
... AT will tell you there is a longdesc available

RS: If you point to details/summary, should provide navigation to it
... We need something to indicate that we don't stringify and that it just points at something
... There is two ways to represent this visually
... You have a link to a site
... or a details/summary
... the latter can include about anything
... the user that doesn't see needs to know that there is a describedby

NOTICE: We're through the discussion time for this topic

<mck> Matt King joined on phone

<ShaneM> The target should be a URI

<ShaneM> an IDREF is just a degenerate case of a URI

TS: Can't go back to DPUB with this

JS: We're not deciding today.

<clown> Suggested reading: http://www.w3.org/TR/wai-aria-practices/#Descriptions_external

@clown any reason that can't be used?

<clown> MichielBijl: no, since it has been a recommended "best practice" for a while.

<clown> at least, I don't see why it couldn't be used.

Than what the hell are we discussing?

* long discussion about longdesc *

<clown> http://www.w3.org/TR/wai-aria-practices/#Descriptions_external

CS: that is nice, you don't have to do anything special
... AT should set up something that indicates it points at a link

<clown> "It is not good practice to use the above pattern when the describing element—the <a> tag with @id='longdesc1'—is http://www.w3.org/TR/2011/CR-wai-aria-20110118/terms#def_hidden, since there is no way for a user to navigate to and activate the link. Use the technique only when the description is not hidden. "

<Zakim> clown, you wanted to point out historical use of aria-desribedby in ARIA best practices.

You can tab to hidden links the same way we use skip links today.

I do want to stress that I agree this should be visible to everybody

What kind of books are we talking about? E-books right? People don't expect the same as with a paper book right? Do they mind extra information? Do they mind the links?

RS: Can we drop describedat?

JS: Not until we have a replacement

<ShaneM> Should flow it out using PlantUML

<ShaneM> bad question: what happens when a describedby points to an IDREF that is NOT a well-defined region?

<MichaelC> ARIA wiki: http://www.w3.org/WAI/ARIA/wiki/

<janina> scribe: janina

Action 1736: aria-primaryaction Move to ARIA 2.0?

aria-primary action


<trackbot> action-1736 -- Michiel Bijl to Create an attribute to indicate a button is the primary button -- due 2015-12-10 -- OPEN

<trackbot> https://www.w3.org/WAI/PF/Group/track/actions/1736

rs: No problem by me, just one more thing to test.
... Any objections?

mk: We agreed last week, just wanted to make sue Rich was on board.

RESOLUTION: Move Issue-624 to ARIA-2.0

Coordinating Testable Statements

rs: How do we avoid clashing ...

mc: Probably we need to generate most tests manually, regretably
... Looking for exit criteria ...

<MichaelC> https://rawgit.com/w3c/aria/CR-pub/aria/aria.html

mc: This is the Status statement for the CR publication of 1.1, when we get to that point

[mc walking through status section]

<clown> https://rawgit.com/w3c/aria/CR-pub/aria/aria.html#sotd

mc: Believe grid tests can come over from 1.0
... then apply to role=table
... plus cross referencing grid and table ...

<MichaelC> https://www.w3.org/WAI/PF/testharness/

mc: Noting Test Harness training is Friday at 1PM Boston
... Believe ARIA tests will all come under Core-AAM 1.1

<MichaelC> https://www.w3.org/WAI/PF/testharness/testcases?testsuite_id=4

<MichaelC> https://www.w3.org/WAI/PF/testharness/testcases/add?testsuite_id=4

mc: Most important are testable statement and expected result. Can be broken out by OS
... Will be a base URI for all test files

<MichaelC> Base for all test files: https://github.com/w3c/aria/tree/master/testfiles/1.1

mc: Noting all tests are migrated away from hg
... Expect tests will be on github, but don't have to be

rs: so git pull will get all tests?

mc yes

rs: Subdirectories?

mc: Think people creating should organize as they see fit

<MichaelC> https://github.com/w3c/aria/tree/master/testfiles/1.0

mc: Only point of directories is to group related tests

rs: Will you migrate testable statements?

mc: Can. Need to know which
... Expect most will be new. We don't need to redo 1.0 tests, only new stuff
... So will migrated on an "as needed" basis

rs: Who's working the next few weeks? Anyone?

fe: Me

ts: Also me

clown: not me
... no commitment--but partially tracking
... grid test files were incorect in 1.0

mk: Especially as we're adding table and mixed grid/table

cs: Maybe similar to tests I put up for table with children

rs: 2 rows and 2 columns as a minimum?

clown: 3x3 better

mc yes, 3x3

mc: to cover not first or last

cs: Need to think about where grid could break and test for that

clown: same with table

cs: Also interaction between html table and aria table
... we have this problem with lists, logical to believe it will happen with table

tracker migration

mc: Found a few issues closed but with open actions
... May be additional croft
... Can people look at their own and close things we don't need to migrate?
... Also pftracker is member only, but new tracker will be public
... We need to know if anything shouldn't migrate for that reason

rs: but don't touch the tracker otherwise?

mc: No, ok to use
... Issues will begin somewhere around 2k

<MichaelC> ARIA Tracker

<MichaelC> APA Tracker

mc: All items will have same id
... including action numbers

<MichaelC> SVG A11Y TF Tracker

mc: Bindings to irc will change in January

e-eggnog all around ...

Summary of Action Items

Summary of Resolutions

  1. Move Issue-624 to ARIA-2.0
[End of minutes]

Minutes formatted by David Booth's scribe.perl version 1.144 (CVS log)
$Date: 2015/12/17 19:01:02 $

Scribe.perl diagnostic output

[Delete this section before finalizing the minutes.]
This is scribe.perl Revision: 1.144  of Date: 2015/11/17 08:39:34  
Check for newer version at http://dev.w3.org/cvsweb/~checkout~/2002/scribe/

Guessing input format: RRSAgent_Text_Format (score 1.00)

Succeeded: s/pint/point/
Succeeded: s/brokinn/broken/
Found Scribe: MichielBijl
Inferring ScribeNick: MichielBijl
Found Scribe: janina
Inferring ScribeNick: janina
Scribes: MichielBijl, janina
ScribeNicks: MichielBijl, janina
Default Present: janina, Michiel_Bijl, TzviyaSiegman, fesch, Rich, MichielBijl, MichaelC, George_Kerscher, JemmaJaEunKU, ShaneM, LiamQuin
Present: George_Kerscher JemmaJaEunKU LiamQuin MichaelC MichielBijl Rich ShaneM TzviyaSiegman fesch janina
Regrets: BryanGaraventa JonGunderson LeonieWatson JoanmarieDiggs
Found Date: 17 Dec 2015
Guessing minutes URL: http://www.w3.org/2015/12/17-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]