25 Jul 2002 - WCAG WG Teleconference Minutes


Avi, Jason, Matt, Wendy, Bengt, Loretta, John, Paul, Lee


Eugenia, Ben, Doyle, Gregg

Action Items

Questions or updates from F2F last week

None. Updates on action items.

Requesting review of WCAG 2.0 TR WD (once published)

AA Society for Technical Communication? Have localization group. There is a usability interest group.

JS The usability of the document itself. Is it providing good technical communication.

AA IEEE communicators group.

WAC IEEE had been writing accessibility guidelines, perhaps review and/or adopt these?

AA Professor in tech comm dept.

JS Rhetoric division here (UT).

WAC Would like more review by disability organizations and advocates.

JS Jay Leaventhal at AFB. Could find some reviewers for people who represent people who are deaf. Gallaudet. National Org on Disability?

JW There are people around here. I will ask them to send comments. OZeWAI-related folks.

JS Ed Bosson. Will check with him.

WAC Loretta, internal to Adobe, what sort of review might happen?

LGR Could try to do something. I'll talk with Doug about this.

JS Jerry Johnson - texas' accessiblity guidelines. state level policy stand-point.

JW policy-setting bodies around the world.

WAC WebAIM class where can get reviewer from devs?

PB Can throw it out to people in workshops.

Process review

JW Process for next few weeks: dealing with issues related to doc, techniques, dealing with comments that come.

JS If a 2 month comment period for review, do we continue working on own refinements?

JW We keep working.

WAC We will have the stable TR for the 2 month review period, but we are likely to have several internal incremental working drafts.

JS What is the formal meaning of last call?

WAC perhaps a bad analogy: but we are stepping up to the plate hoping to hit a run to get us to CR and then advance to PR and Rec. Last Call means very stable. We feel we've dealt with major issues and are ready to progress to CR. Serious step.

Issue #674

JS Issues with 1.0.

WAC right, 1.0. No one is trying to adopt 2.0 yet since a WD. EO doing interesting meetings, 2 x's a week. Subgroups beneficial. A meeting to accomodate Europe. Perhaps once a month to check in.

JS Africa?

WAC No. No one in Africa, S. America.

LGR India?

AA We have 52 languages represented here. Perhaps could scare up some contacts.

JS UN Virtual volunteers program. They work in diff locations on projects that might be far away from them.

MM WAI has contacts in SIDAR and organizations in Spain and France. Spanish speaking countries in Latin American could have connections, similar to France and Africa.

JW Gets at review issue. People who review might get interested and involved. Nice to get people in expertise in SVG and other formats. Some happened last week. People who have time to write techniques and prepare them for inclusion.

JS Also on internationalization (reviewing and techniques) solicit examples of good and bad sites from people in other countries.

WAC CR exit criteria. Likely to need sites in a variety of languages.

JS Especially Hebrew.

WAC right, in relation to readability in braille.

JW Editing is not complete. The language could be fixed. This draft is not necessarily idea of reviewing the language.

JS Need to make sure that applies to other languages.

JW Sending issues for next week. One of them, checkpoint 5.2.

WAC what about 4.1 and 4.2? AA, BF, LS, and me try to get together to discuss (finish discussion from F2F). Invite others as well.

AA I'm meeting tomorrow with Anetta Cheek to discuss. Other questions?

JW Primary questions are about success criteria. They need to be testable. One possible success criteria "this content has been reviewed by others and found to be as good as can be..." That is a testable criteria and can take into account the non-testable items.

AA One on the list was "avoid splitting infinitives." That is fairly controversial.

JW A matter of adding to our list of strategies and thinking about success criteria.

WAC Issue with use of "site" throughout. Can't always apply to "app."

JW Should probably use "content."

WAC One issue that came up last week often was baseline browser capabilities.

JW We addressed it in checkpoint 5.2, that the author can set their own baseline. Agreed to include sample baselines. Part of core techniques should deal with how to set a UA baseline.

LR How far backwards compatible will we be?

JW The idea of allowing people to set their own baseline was to let them choose with our guidance. With Level 2, the proposals were to lay down general requirements. e.g. at level 2 the technologies in the baseline have been available and internationalized and compatible with assitive tech for some period of time. In the guidelines document we can't set down any standard of backwards compatibility, since what is appropriate will change over time.

WAC Was anything drafted for that?

JW Just an idea that was proposed, nothing drafted. Gian had an action to draft level 2 success criteria. No one has taken action to work on techniques for it. I can draft level 2 success criteria. Discuss

WAC Anyone interested in drafting something?

LR One version prior to the current version.

JW The main issue is that it doesn't specify which technologies.

LR We're talking about UAs?

JW You mean for every UA that supports X, that UA must have supported it for a previous version?

LR If lay out all of the standards and realize that IE 5.5 valid version current is 6. when they go to 7, then 6 is the current.

WAC how details? platforms, UAs, format spec. e.g. "CSS1, IE5, NS4, etc etc"

JW Not specify particular UAs. "Designed under assumption that UA supports HTML 4.01, DOM Level 1, CSS 1..."

WAC What about IE and NN that claim HTML 4.01, but not every element or attribute supported (i.e. longdesc) or they don't suppor the same way.

JW Then specify which features you are not depending on. They could be written in text. Currently no restriction on how you do it. Also possible to use metadata to specify.

JS If it gets to the point where the dev has to say, "our baseline is HTML 4.01 except for longdesc, abbr, etc. that we no one supports" that seems attenuated on someone reading the claim and too much work for dev.

JW How would you specify differently?

JS I'm more accustomed to specific browser versions, like Lee said.

JW Difficulty if in our doc we specify based on specific implementations. Implementors not listed would be upset.

WAC As long as not an edorsement and an example, i think we are ok.

JS We have validated against HTML 4.01 transitional, if there happened to be browser support issues that is the whole world's problem. It's everyone's issue.

JW 5.2 says "features above baseline...content should still operate." Perhaps not qualify baseline much. If decide not to use certain features of standard, you might want to document internally but why appear in baseline if for purposes of checkpoint things that aren't supported won't be used.

JS No requirement to use longdesc for valid HTML. If you have a rich image, you can make that description part of page content. longdesc is a judgement call.

JW Given our examples, list the standards you are planning to use.

WAC what about scripting and web applications?

JW Include script support in baseline. If it is part of the baseline, then it must be supported for the content to work.

JS The end user has no basis for complaint if have scripting turned off?

JW If the dev decides scripting is in baseline then they are conformant to level 1 (as currently written). They might not be able to reach level 2, depending on how we define it.

JS Want to recommend, likely a techniques thing, that there be a front page to such an application that says, "if browser does not support the baseline, you're out of luck?" Would we recommend that devs identify up front those apps that will fail if have scripting turned off?

LR We'd end up w/problems. People still using jscript and only programming for IE. What if we say "they must comply to w3c standards." jscript not in that.

WAC We provide the DOM and say use the language that you want. Plus, point of W3C tech is to extend (XML, RDF) thus people should be able to use other langs.

LR Issue with third-party plug-ins.

JS If use, they should be accessible.

Action LR, WAC come up with baseline capabilities proposal. first step dig up previous proposals, then think about how to improve upon.

Action JW write level 2 success criteria for checkpoint 5.2

$Date: 2002/07/25 21:38:23 $ Wendy Chisholm