User Agent Accessibility Guidelines Working Group Teleconference -- 12 Aug 2010

<trackbot> Date: 12 August 2010

<Jan> ok will try again

<kford> in a word no, just abig oloud buzz.

<jeanne> let me know when it stops. I can't listen to it.

<jeanne> scribe: jeanne

F2F in Leon, France (go/nogo)

JA: Last May many said they could do it, maybe. Now it is time for plane reservations.

GL: I will be in FLorida, so it would have to be the later part of the week.

<Jan> TPAC=Nov1-5

UAAG meets Thursday-Friday

JR: I can't make it in person

<Jan> JR: Can't make it in person

JA: Simon will be in the US for a conference

<AllanJ> discussion continues, video conference, different location

<AllanJ> KF: last weeks video conference call was very useful

KF: For those who want to be the representative of user agent, it would be useful to go, but for most people it is not.

JA: But we don't get the joy of breakfast, lunch and dinner together. :)

KF: I am not convinced that going to France to do that is the best use of time and financial resources.

JS: It is ironic that our European members are the ones that can't attend. We wanted to have it at TPAC because it was easier for the EU members.

KF: I cannot attend on my own money, because Microsoft won't allow it.
... HTML is too big, and we can't interface with them directly. We have to go through PF and A11y Taskforce, so that is not a good reason. There are serendipitous conversations, I don't mean to discvount that, but I don't think it is worth it.

Resolution: UAAG will not attend TPAC and meet there.

<scribe> ACTION: Jeanne to contact W3C admin to change our meeting. [recorded in http://www.w3.org/2010/08/12-ua-minutes.html#action01]

<trackbot> Created ACTION-424 - Contact W3C admin to change our meeting. [on Jeanne Spellman - due 2010-08-19].

<kford> Scribe: kford

Longdesc

JA: Issue around longdesc came up at wai CG meeting.
... HTML5 working group has saidd we are not going to support longdesc in html5.

<Jan> HTML5 Decision: http://lists.w3.org/Archives/Public/public-html/2010Aug/att-0112/issue-30-decision.html

JA: Said current laws and specifications were not valid reasons for including.

<Jan> WAICG Resolution: http://www.w3.org/2009/06/Text-Alternatives-in-HTML5.html

<jeanne> close action-424

JA: There an an effort to have this reversed or file a formal objection.

<trackbot> ACTION-424 Contact W3C admin to change our meeting. closed

GL: What was the reasoning saying this was harmful?

<jeanne> http://lists.w3.org/Archives/Public/public-html/2010Aug/att-0112/issue-30-decision.html

JR: They think it hasn't been used and no good use cases.

<jeanne> JA: They said WCAG, 508 and other international law was not sufficient reason.

GL: When I've tried to use longdesc I've been a bit frustrated by the syntax.

<jeanne> GL: I am frustrated by the syntax they showed - requiring a separate page rather than having the option of having it in the same page.

JR: Good point, there are problems with longdesc. ARIA DescribedBy is an attempt around this.

<jeanne> JR: People recognize that Longdesc is flawed. But PF don't want it Obsoleted. They want deprecated by still conforming.

JR: PF is concerns about the jump to straight non-conforming, prefer deprication first.

<jeanne> http://lists.w3.org/Archives/Public/public-html/2010Aug/att-0112/issue-30-decision.html

Group brings Kim current on discussion.

<jeanne> Current Chairs are: Sam Ruby, IBM,

<jeanne> Paul Cotton, Microsoft,

<jeanne> Maciej Stachowiak, Apple

<jeanne> KP: there should be a converter that automatically changes longdesc to ARIA describedby.

<jeanne> GL: How much is it used? Can google get data?

Group continues to discuss longdesc.

<jeanne> JS: that is how it has been justified in the past - because use is low, that justifies removing it. But it ignores all the use cases that are behind firewalls, like financial houses that use it for bar and pie charts are behind personal passwords of their clients.

<jeanne> KF: The real problem is the obsolete status for all the companies that have done the right thing, it penalizes them for putting accessibility in, by making their web pages invalid.

<jeanne> JA: We won't solve it here, but we need to decide what, if anything, we should do as a group.

JA: Four us as a group to we want to do our own thing, join any formal exception or let this pass?

<jeanne> JS: I think we should get our employers to object.

<jeanne> KF: I think we should object as a working group.

<jeanne> KF: the way this was written in HTML4, any browser that didn't implement the technology doesn't conform to UAAG 1

<jeanne> GL: the UAAG 2 says "implement the technology as it is spec'd"

<jeanne> JR: UAAG 1.0 has Level A - support accessibility features. Level AA - implement the spec.

<jeanne> GL: I think this is not directly our issue - PF is really the group that should be working on it.

<jeanne> KF: If there is an objection filed and a way for UAAG to sign on to that, I would like to do that.

<jeanne> JR: longdesc is still a valid attribute of HTML4. What do you think browsers will do in the meantime?

<jeanne> KF: It will still be in the DOM. I don't see that changing.

<jeanne> JR: It is a political issue in how HTML5 deals with PF, but is it a real issue, if that is the way that the browsers will deal with it.

<jeanne> KF: I think it is a big deal in the way that we as WAI representatives were dealt with by HTML5.

<jeanne> JR: It is not a good idea for User agents to have large backward compatibility break.

<jeanne> KF: So there should be a way to make it better. Is there a way to get resources together to make sure that accessibility is taken care of in some way.

<jeanne> JR: If HTML5 is going to break backward compatibility, then they should bring in ARIA natively.

<jeanne> KF: If we have the opportunity to sign on to a formal objection, who would agree to that?

<jeanne> KP: and I would like to see a solution to convert people to ARIA describedby.

<jeanne> JR: UAAG is concerned about the loss of backward compatibility whenever is concerns the ability of a browser to bring an accessibility feature to its users.

<jeanne> KF: Even though the general browser support of longdesc is poor, the assistive technology companies have implemenated longdesc.

<jeanne> ... Don't throw out the old technology, you can discourage its use, but don't throw it out.

<jeanne> KP: if something is different, it is more difficult. There are accessibility issues with slow - tools are slow and they don't have to be.

<jeanne> ... the net neutrality is a related issue -- that sites that people rely on for accessibility are more likely to land in the slow lane. So people who need accessibility will be worse off.

<jeanne> JR: Every major platform has accessibility APIs supporting longdesc. Why would that engineering effort be made if it wasn't useful.

<jeanne> ... describedby has it's own flaws.

<jeanne> JS: Would like to suggest postponing this discussion until we hear what PF will do and what John Foliot's proposal look like

<jeanne> [group agrees}

Writers Meeting Survey http://www.w3.org/2002/09/wbs/36791/20100802/

<jeanne> scribe: jeanne

Delete 1.1 and 1.2

<AllanJ> group discussion at writers meeting http://lists.w3.org/Archives/Public/w3c-wai-ua/2010JulSep/0030.html

JR: Color contrast is not an accessibility feature the way longdesc is.
... the whole interface between what the tool does and what it renders are different standards.
... the user interface of the tool must have good color contrast, but if the author coded the content with poor colors, then there isn't a lot the user agent can do.
... images of text is a better example, because the user agent can't fix that.

JA: So this is a compelling case to keep 1.1. We keep the 3 levels for WCAG, but at least we simplified the non-web interface requirements.

<AllanJ> http://www.w3.org/TR/ATAG20/#principle_a1

<Jan> In ATAG2 they are in the revewrsed order

<AllanJ> A.1.2.1 Non-Web-Based Accessible: Non-web-based authoring tool user interfaces follow accessibility standards and/or platform conventions that support accessibility. (Level A) [Implementing A.1.2.1] Note: If a conformance claim is made, the claim cites the accessibility standards and/or platform conventions that were followed.

<kford> Group has extensive discussion on this.

<AllanJ> JA: 1.1.1 seems to say the developer of the UA should not do inaccessible things, and document it

<AllanJ> ... the documenting part is covered in 5.3.

<AllanJ> ... the 'not doing inaccessible things' is what is important.

<Greg> Perhaps combining the sub-items using bullets, such as: 1.2 WCAG Compliant: User agent user interfaces that are rendered using Web standard technologies conform to WCAG: • User agents claiming conformance to UAAG Level "A" conform with WCAG Level “A”. • User agents claiming conformance to UAAG Level "AA" conform with WCAG Level “AA”. • User agents claiming conformance to UAAG...

<Greg> ...Level "AAA"...

<Greg> ...conform with WCAG Level “AAA”.

<AllanJ> JR: if use same approach as ATAG, should use similar wording. Plea for review of ATAG last call

<AllanJ> ACTION: AllanJ to rewrite intent for 1.1.x include Jan's example. [recorded in http://www.w3.org/2010/08/12-ua-minutes.html#action02]

<trackbot> Sorry, couldn't find user - AllanJ

<AllanJ> ACTION: JAllan to rewrite intent for 1.1.x include Jan's example. [recorded in http://www.w3.org/2010/08/12-ua-minutes.html#action03]

<trackbot> Created ACTION-425 - Rewrite intent for 1.1.x include Jan's example. [on Jim Allan - due 2010-08-19].

<kford> who is closing i t out?

- DRAFT -

User Agent Accessibility Guidelines Working Group Teleconference

12 Aug 2010

Attendees

Contents

F2F in Leon, France (go/nogo)

Longdesc

Writers Meeting Survey http://www.w3.org/2002/09/wbs/36791/20100802/

Delete 1.1 and 1.2

Summary of Action Items

Scribe.perl diagnostic output