User Agent Accessibility Guidelines Working Group Teleconference

17 Jan 2013

See also: IRC log


Jim_Allan, Jeanne, Greg_Lowney, sharper, Jan, Kim_Patch, Markku
Kelly, Kim
JimAllan, kellyford


<trackbot> Date: 17 January 2013

conformance use cases for conformance scenarios (extensions, mobile apps, etc)


<Jan> GL: in 2nd one, UAWG should be UAAG

<scribe> scribe: Jim

gl: in item 2 - confused by 'all functionality'

<Jan> (2) the user agent COMPONENT (plug-in, etc.) must meet any requirements applying to all functionality (e.g. to be resizable, to provide documentation, etc.).

jr: right, can't hand off everything. If the guidelines say everything must be documented, then everything must be documented even for partial conformance

gl: trying to be inclusive

jr: COMPONENT, need to insert into 2.

This conformance option may be selected when user agent (or plug-in, extension, etc.)

would be This conformance option may be selected when user agent component (eg. plug-in, extension, etc.)

gl: this would be one that nearly all browsers would use, because they will require at least 1 extension to meet all UAAG requirements
... so using term, partial conformance, a browser may not comply except with an add-on, or partial conformance could only be for an add-on
... take 2 browsers. neith comlies out of the box.
... FF has a plugin that allows it to fully comply. BrowserX doesn't have a plugin so it does partial.

jr: FF could do partial, but say we have all the archetectural features to allow for a plugin. if X doesn't have the architecture, then it could not fully comply

js: add 3rd example - does not meet 2.2.3 and does not have extensible architecture then they could not claim conformance

jr: had pushback with ATAG. folks would not make a full claim with extensions, because the extension maker could make case for payment for use of extension

gl: need some language somewhere to make it clear that person making the claim, when and how they claim full copliance when using 3rd party add on. versus partial - where their architecture allows (theoretically) the use of adons.

<Greg> Want to make sure the document makes it very clear how the person creating the claim when and how they claim full compliance relying on third-party, optional components (e.g. the mouseless browsing add-on) vs. partial compliance in that their architecture in theory allows creation of such an add-on but either it doesn't exist yet or they don't want to be reliant upon a specific available add-on.

js: don't like idea of partial conformance of some theoretical existence of a plugin.

jr: was written specifically that way. theoretical possibility that something could be written. it shouldn't require you to name names of plugins (that would be full compliance)

gl: this would allow any open source project to claim full compliance- because the source code is available and anyone can write an extension

jr: that seems far fetched

gl: what is the reasoning again...

jr: when you make a claim, open source browser, it does not do mouseless browsing, but it has an open architecture that allows plugins that get keystrokes, insert content.
... vs other product, that doesn't have an extensible architecture, if you want mouseless, just rewrite the source.
... could be another type of partial conformance. there should be a way to make a claim without naming names.

js: I get this.

jr: mouseless browsing, partial conformance. have their relationship with FF. mouseless should not care that FF can't highlight words. they are only about mouseless
... what if there is a bunch of mouseless browsing plugins. FF could say yes there are mouseless plugins but not saying names.

<Jan> http://www.w3.org/TR/WCAG/#conformance-partial

kp: the example should exist, and give examples (x, y, z) or including blah
... 2 different things...I do 1 thing. and I do everything except for x. there are extensions that meet X examples are foo, foobar, and fubb

gl: but if there is a browser with a closed architecture, with a specific API, and there is only one extension. no-one is able to write others, etc.

kp: it must be documented.
... difference between open/closed architecture is the documentation

jr: extension exists, do we have to name them.
... host browser, definitional split. mouseless doesn't have to meet longdesc, and a browser ???
... extension as a browser. PDF is an extension. but is also a browser

<Jan> JR: Its writtenh the way it is because splitting user agent from plugins is too hairy

<Jan> JR: ie a plugin could be a user agent

<Jan> JR: so instead we just call it partial conformance...

<Jan> JR: and if a full browser wants to claim its a component...fine...its a lot of work to do...just to then claim you are only a component

js: good here, need some tweekes (editorial?), need a decision tree.
... full compliance must name names of extensions

partial- conformance: theoretical possibility of an extension, or that architecture allows extension, and there are some in the wild

scribe: there must be documentation as to the extensibility

ja: distinguish between browser missing a few bits, and an extension which is 1 bit.

gl: plugin as UA - pdf, media player,

jr: PDF is both.
... could imagine a tool that sends everything to a rendering engine (transcoder) that is outside of itself.
... water is muddy. software that hands things off vs doing something. mouseless does something, it is the end of the line... the browser hands things off.
... browser is an onion, it does a lot by itself, but can't do somethings so it hands it off.
... mouseless takes the handoff, does its thing, and gives it back to the browser

<Jan> JR: The concept of "hand offs" could be useful

jr: don't want loopholes

mh: partially compliant, with a theoretical extension. does the 3rd party add on interact with platform API or through a custom API

gl: extension could communicate directly the the a11y API or communicates through the extension architecture of the browser and the browser chats with the a11y API
... pdf viewer plugin

<sharper> While I think it is useful to make the best job of this as we can. It seems to me that we have to be working in good faith. These companies will have Lawyers, paid millions of [pick your currency] who would probably be able to run rings around anything we write here if they want to. I'm sure there will be many loopholes in combination with law, and companies will be able to get around lots of stuff if they want to.

Feb 8, another meeting to find apps and extensions that meet SC.

Summary of Action Items

[End of minutes]

Minutes formatted by David Booth's scribe.perl version 1.137 (CVS log)
$Date: 2013-01-17 19:39:38 $

Scribe.perl diagnostic output

[Delete this section before finalizing the minutes.]
This is scribe.perl Revision: 1.137  of Date: 2012/09/20 20:19:01  
Check for newer version at http://dev.w3.org/cvsweb/~checkout~/2002/scribe/

Guessing input format: RRSAgent_Text_Format (score 1.00)

Succeeded: s/definiitional/definitional/
Succeeded: s/APIO/API/
No ScribeNick specified.  Guessing ScribeNick: allanj
Found Scribe: Jim
Default Present: Jim_Allan, Jeanne, Greg_Lowney, sharper, Jan, Kim_Patch, Markku
Present: Jim_Allan Jeanne Greg_Lowney sharper Jan Kim_Patch Markku
Regrets: Kelly Kim
Found Date: 17 Jan 2013
Guessing minutes URL: http://www.w3.org/2013/01/17-ua-minutes.html
People with action items: 

[End of scribe.perl diagnostic output]