24 August 2000 WCAG WG telecon

Summary of action items and resolutions



7.4 Errata

WC Summarizes issue.

KHS Yes, we should look at.

Marshall: REfreshing fine, if we require that it starts out static and let the user set up refresh if that's what they want to see. That is strict, but the user can control speed.

CS Perhaps say "meta refresh" tag in static HTML file where not generated on the server based on user settings is a problem.

WC Interesting, but changes clause.

JW It is an until user agents clause, one way to compensate is to provide a form and ensure the updating is switched off or that the time period is set by the answer. Therefore a way to satisfy the checkpoint.

Action WC: Rewrite to say that as long as user can stop or control the rate (through an author provided mechanism) is ok. Say something about use of META in static HTML file.

Action CS/WC: Create examples to show this would be done and then document in Techniques document.

?? Default be static then let user change if needed.

Marshall: if they want to set preferences can be stored in cookie.

WL Raggett and index doesn't say anything about META.

/* it's a hack */

JW Even tho not in standards.

Matt it's in HTTP spec.

WL Should we also be addressing protocols and not just markup languages?

WC Haven't forgotten about, and not excluded from WCAG. Discuss very vaguely, need details to put in techniques document.

WL Anyone working on in W3C?

WC Yes. HTTP handed to IETF, CC/PP current work.

Action WC: Document as open issue for techniques: protocols and server-side issues.


WC will publish new draft tomorrow with editorial changes.

CS What about my suggestions?

WC More substantial, next version (not tomorrow). Concerned about making too many changes or additions after IG/WG thorough review.

Technology-specific checkpoints

WC Concerned that we're writing the abstract before the document, whereas I think I was taught that one writes the paper and then summarizes it in the abstract. Perhaps we need to start in the technology-specifics and then generalize.

WL Think we can do in parallel.

JW I made sure that original checkpoints fit under the new draft.

CS Comments on "use style languages correctly" it seems that HTML table checkpoints would fit in that. E.g., avoid using tables to make things look pretty.

JW Table is structural. Or fit under "user markup according to specification."

CS "Where presentation is used" that would apply to use tables for tabular data, to mark up columns and rows.

JW Hesitent about putting tables for layout under 2.2, but not obviously wrong, fit better under "use markup correctly."

WC /* tries to understand the difference between all of the current guidelines under principle 2. doesn't see that it is clear that structure and semantics ought to be capture in markup */

JW It's not sufficient to say, "capture structural and semantic distinctions" but must say which ones to capture. Additionally, those in presentation need to capture as well.

WC Trying to distinguish between presentation semantics and structural semantics?

JW Whatever is in presentation has to be in structure and semantics as well. Related: in any event there are distinctions that need to be expressed in markup. Then depends on how content is going to be used.

WL /* reads 2.3: Where presentation is used to communicate distinctions of meaning or structure within the content, ensure that the meaning is captured in the markup. */ Presentation is almost always used to communicate distinctions. Pure eye-candy is very rare. Presentation is almost always used to communicate.

DB There are different levels of what an author is trying to communicate. Agree with William to a point.

WL This is something we have to deal with.

JW We are discussing requirements regarding structure and semantics, more generally we are considering whether parts of these need to be clarified to determine if technology-specifics will fit under them. /* opens the floor for others */

Matt The statement could be too specific to HTML. In Flash or SVG there may not be a way to adequately extract presentation from semantics. There may be something in this specific phrasing that would preclude other technologies from being accessible.

JW CMN isn't here, he has been arguing that there is higher level of semantics that can be captures in SVG than in other graphics formats.He has various ideas for using metadat to provide additional semantics to complement presentation and grouping with semantics provided by RDF.

Matt My issue is that we have a well-defined separation between semantics (XML, XSL, etc). that gets muddier from standard to standard being able to clearly delineate X is semantics, Y is presentation. It's more an issue of using elements of auditory, tactile, or visual design so that they can not be accessed by other means. By saying "it has to be semantic or presenation layer" is not possible to achieve in certain formats. However, "don't use elements of visual design to communicate semantics" - something more specific to the overall goal that makes it accessible to all media.

JW I personally disagree that it become increasingly difficult.

WC Eye-tracking results...

/* discussion about how the eye follows visual distinctions on the page */

DB What are we trying to do?

JW Regarding the requirements under Principle 2 we could reorder them so that those that demand structure are in one place. Then a matter of clarifications and rewording as needed.

WL Think in good shape, but need examples. Hard to know what they mean.

WC Think we can say them clearly without lots of supporting material.

JW I will volunteer to take an action item, what do people want me to do?

WL Need an introduction. Why are we harping on structure so much? We want people to play on a level playing field.

WC Propose that WL write intro and JW work on principle 2 guidelines to provide explanations you think people need to understand.

JW Ok, what do people want me to do?

KHS Make it understandable to the general population.

Action WLWrite a piece about structure, presentation, different media, orientational and navigational issues in reading content.

WC KHS do you have ideas for how to make it easier to understand?

Action KHS: Send suggestions on how to make the document easier to understand.

Marshall: The average Web designer may know that size attracts, but many do not think about the effects of flashing text or big text or how it will affect a sighted person much less accessibility issues. We might suggest how a person with sight will use the sight.

WL Until you find that a machine can go through a document and create a table of contents, for people who have not dealt with it's a miracle.

CS Small sites: why automate it, can build by hand.

WC In regard to Marshall's comment, addressing the visual presentation addresses needs of people with cognitive and learning disabilities and thus is very much in scope.

Action JW: Editorial fixes on text of guidelines.

Marshall I have some design background.

JW Would you like to look at principle 4? Requirements that are not included there that you would like to be there?

Action Marshall: write up something about visual presentation. Look at principle 4 for requirements that are not included.

WL JW, while you're at it, please define transform gracefully.

KHS It makes sense, whether well defined in industry or not. I'm going to look at the document as a "regular joe" and send the comments to the list.

WL Transform gracefully means that things can disappear if browser doesn't support.

KHS Changing from one aspect to the next w/out being confusing.

WL When have CSS and have a device that doesn't support it, it still works. It has nothing to do with transforming.

CS Yes it does. It is often done with server-side transformations.

WL Transform means, it changes from one form into another form.

Marti: are you saying, eliminate?

WL Yes.

/* Charles joins */

/* JW summarizes SVG discussion */

CMN Even a bitmap can carry semantic info. One way to provide info about SVG is to add metadata into it. You can do that in SVG and any commonly used graphics format. There is a Note about putting metadata into JPEG. Can do same thing for gifs and pngs. Most formats allow you to attach metadata in some way.

JW There was an issue if in graphics formats it was possible to distinguish between semantics and presentation.

CMN In SVG where you describe semantics and presentation in XML there are good ways to describe semantics in a way familar to HTML writers. In other formats (GIF or JPEG) where presentation is usually all that is given, you can provide the info but people usually don't do it. Describing and retrieving photos using RDF and HTTP

Action CMN: techniques for wcag2.0 draft for SVG. Based on Accessibility Features of SVG.

WC What about 1.0?

CMN Rather 2.0, then figure out how to do it for 1.0.

Next meeting

Resolved: no meeting next week.

CMN Meet every other week?

WL Too much going on, can't do. Except this once.

CMN Mostly comes from the fact that we've been skipping meeting fairly frequently perhaps we should formalize this.

$Date: 2000/11/08 08:27:09 $ Wendy Chisholm