23 March 2002 WCAG WG F2F Meeting



WAC What are you all interested in discussing this a.m.? What issues are you worried about?

KHS - Status of PDF, glossary

KC Policy, regulation and harmonization


JS Finding out more about the group and process

HB Glossary, excited to see where Daisy is heading. isues w/E-book

MM HTML, want to collect issues to take back to seattle group, Flash

JA HTML, particularly testing


BC Authoring perspective, more familiar w/the group

BR here representing UW research and Macromedia. Therefore, can proprietary languages conform to WCAG? If not, what case can I build internally?

CS /* missed */

GV /* missed */

PDF Techniques

KHS A lot of effort invested, but now where do we take it?

CS Most of the world uses proprietary languages.

GV I'll take off-line and do some research.

HB Is it a document that Adobe would release?

KHS Supposed to be W3C WCAG techniques.

LGR This is an issue to discuss.

CS Member companies do this with other technologies. People submit work to the W3C all the time.

KHS Can we take a look at them later?

GV The structure would be based on WCAG 2.0. Flash - how long do you want to talk for?

BR As long as needed. How do these guidelines apply to web applications?

GV Let's do it as we discuss the new version.


KC Primarily digital rights question.

/* scribe missed most of discussion */

GV Anything else we need to discuss today? no, ok.


GV HTML and glossary

HB What accomplish w/HTML?


JS Merging around web applications. Do we have a framework?

GV This new draft is an attempt to resolve a whole bunch of issues. If it creates issues, then we can revert, go back, etc. If any changes that have been suggested, didn't get picked up, let us know. /* reads from draft */

Notes from HTML Techniques sub-group

Paul, Ben, Matt, Jenae, John, Bob, Loretta, Cynthia, Wendy


summary attribute

applying scope vs headers

metadata - context and orientation

Core techniques

currently in html, could be in core


Other topics


Notes from Glossary sub-group

Katie, Harvey, Kevin

/* on the way */

PDF Techniques

Judy Brewer joined us for this discussion.

JB W3C usually published info about W3C work not info about company-specific work. At one AC meeting, where the member orgs send a rep. WAI presented a highlight that listed the WCAG Techniques docs. Some companies were concerned. We replied, when we address issues of Web accessibility, our priority is to address W3C technologies, but that we look at other technologies that affect accessibility as well.

JB There was a W3M meeting, it's been a while. If we publish techniques for PDF there would have to be a techniques to articulate the relationship.

JB Does not mean that you can not but neither does it mean that you can. It needs to be discussed in w3c process. Publish a module as with the others. Or a slightly different looking module, but in w3c space. Probably will be public concerns, even with a clear disclaimer. It could appear W3C endorsing the use of the tech in an accessible way.

JB Another approach, some collaborative discussion and have Adobe publish something in their own space. However, could say "developed in conjunction with W3C" or something. Again, need to go through W3M.

JB There is a need to address this. let's get it going and then see what we can do with it. However, don't do all the work if you can't do anything with it.

JB What if you work on the front end of it first, what is the appropriate reltionship? Think through where you think the positioning is. Then send that up the pipeline for feedback.

JB Structurally, Adobe is a W3C member there is a process available to member organizations - submissions. Ordinarily it is commentary on something or technical work to be hosted in collaborative space. PDF is proprietary, so not submitting that, but perhaps submit techniques.

JB I could take a proposal to W3M about what is the best process. These are some interesting issues that I'm sure we would be interested in discussing.

HB Subsets that are proprietary and others that would be public.

LGR The PDF spec is public.

CS Also, how to make an activex control is available.

JB But it is controlled by Adobe. The accessibility transformation hypothetically could be a submission.

KHS The best suggestion is to come up with the right wording.

JB a clear statement of relationship. In the process of writing something up, then bat around the WG, it might become clear what the right thing is to do.

LGR My biggest concern is that people will start with WAI and there should be a path so that they can find the info.

JB Good point. What are the paths?

CS I'm guessing that if i put something on microsoft.com, the style required is different than WAI work.

JB From w3c point of view, there's all sorts of things in the intersection.

GV Helpful if we try to thoroughly try to examine from the various directions.

JB Try to do clarification of placement first, before working more on techniques. However, someone looking at the issue would want to see examples of what will likely be said.

HB Meta model of this to apply for other vendors.

KC Legislation put a requirement of providers of goods and services to make systems accessible, but not each piece has to be equally accessible. I don't midn if has Flash as long as I can get the info. Every kind of product has to reach a certain level of accessibility.

CS Flash and PDF have had work to make accessible and people don't know how to do that. People should know how to use the tools that they want to use. If using an ActiveX control, how do you create a text equivalent? If that system is tied into the OS features and works with screen readers. It is possible, but no documentation.

BR Issues of innovation. I want to make sure that as future versions are released, we address current issues as well as upcoming issues. A chance to push-back a little rather than by saying "you don't exist"

CS Web authors would do accessibility if they could do it and have a cool web site. If the cool stuff is accessible, it will resonate better w/web authors.

PB Writing a doc and have it disappear, it could be housed in a number of places. W3C, Macromedia, WebAIM.

BR I need to have a fair amount of confidence that it will be useful. b/c I am one person, I have to make sure that with my limited amount of time, I am using it wisely. This is one product, we have 14.

Joint meeting with WAI IG

We met with the WAI Interest Group for just over one hour. Gregg gave an update of the progress of the group and the WG members who were present answered several questions about the WG's work.

Summary of questions and answers

Mark Urban: How does WCAG 2.0 address cognitive disabilities?

GV Several ways:

  1. Cognitive disabilities are currently addressed by about 80% of the checkpoints.
  2. We will work with cognitive tools that are coming down the line.
  3. We are providing more materials.
  4. Several of the checkpoints help all, not just cognitive. For example write clear and simply.

Identifying the issues we want to resolve has been harder than solving them.

Marie: Who can become a participant?

GV Someone who listens, has stamina, and who is willing to read. /* describes participation criteria and how to join */

Brian Hardy: Timeline for WCAG 2.0?

JB describes the process that the document must go through (working draft, last call, candidate rec, proposed rec) and that it is difficult to predict. Focus on Candidate Recommendation and getting implementations.

GV We need a bunch of sites to apply it to, to try it, and who will come back to us asking "what exactly do you mean here and here."

/* We left the IG meeting and returned to our own meeting room */

WCAG 2.0 Restructuring proposal

GV It's not a new draft of 2.0 it is a restructuring proposal. If it looks good, we can head in this direction. Each item has the old item in square brackets. We didn't want to equate the name with the content so that people wouldn't think a movie was accessible if you had given it a title.

/* GV describes rewording of checkpoint 1.1 */

JS It sounds like there is an assumption that what can be expressed in words can always be expressed in other forms where there are things in other media that can't be expressed in ... "can't be expressed in other forms" there are things in verbal structures that don't lend themselves to expression in gesture or sound.

KC Give them the score if they can't hear it. Those aren't words. Chess notation.

GV You can express the symphonic score in words. It would not fall outside. "This note followed by ...."

KC It's not words, it's symbolic notation.

GV I can use words to describe your symbolic notation.

/* JW joins */

KC I can live with that.

KHS What is the glossary entry?

KC Have to explain that in notation, could be expressed in words.

JS If you are limiting the scope to finding equivalents for individual words then that's a different thing than finding an equivalent for verbal content.

GV A different category.

JS Important to say that. In 1.0 the assumption is that information is stable and not affected by the form in which it is expressed. That is not a universal assumption.

GV Good point. This just talks about whether you can perceive it, not if you can understand. Just can you tell if it is there. This will be important for tools and access.

KB I like the intent but there is a bit of confusion about what it might mean. example of music. would this create a requirement to express the score in words?

GV No, one guideline says "the visual track of a video has to be presented in words." Not talking about not having. What we require will be in the checkpoints, it's defining perceivable but putting in a caveat that if we want people to do that we need to acknowledge. In the cognitive discussions, some people claimed "but everything is available to people who are blind." That's not true, there is information, such as the Mona Lisa, that is not experiencable by people who are blind. If something can not be expressed in words, we have not touched the symbolic part for expression. If you put up a symphony do you have to put up a score.

CS Assuming you have the rights to do so.

GV Trying to get concepts and what are the checkpoints.

KB Does this become a checkpoint that is dependent on the creativity of the writer? If i were to take a picture and passed it around and asked everyone to say which content could be expressed in words or not....meeting this guideline by identifying the content that could be expressed in words, we would get a variety of descriptions. people will pick up on different details and different levels of details. It might rely too much on the idea of expression on words.

KB Does the concept of expressing in words, is words the right terminology for all languages. Does that work in Chinese?

Action GV: does Chinese, signed languages, etc. have "words." If we use this language, does it imply that the only way to make it perceived in another modality would be to use words instead of notation, e.g. music and score.

CS I really like what you are trying to do. What is the first question that a person would ask when reading this? I tried to do that with rewording on 4.x checkpoints. MusicML and MathML are better representations. Seems that you are trying to define "equivalent" and taht the concept of expressability in words is relevant only to text equivalents and not other kinds.

GV Trying to create a checkpoint where the success criteria required something not in the checkpoint or the flip (the checkpoint says "you must do this" but the guidelines say "only for the following 2 things.")

CS There are things about presentation that might not be expressable in words. It seems that this applies to 1.1 but not perhaps other checkpoints there now. e.g.. using MathML instead of picture of equation.

GV 1.3 says "content and structure..." the fact that presentation can not be expressed in words is a reinforcement...get content out of presentation.

CS If you are using an alternative that is not a text alternative...perceivability doesn't mean you have...

GV They are words.

WAC Does that mean that SVG is words?

JW Replying to KB and CS. Chinese has words, no doubt about that whether spoken or written. Sign language have a grammar, there are units that have a role. The role they play in the grammar of the language.

/* scribe missed 2nd point */

KC We need reference to the authoritative source if there is one. For non-language texts there is one. Also, info about the intermediary (interpreting the image). You need to know not only what it looks like but what it is for.

BR We have questions we have to answer. /* missed all of the point, primarily what are the exceptions. how do i know when this passes...who decides? */

CS The question for 1 is "What about the Mona Lisa?" There is a technical definition of text that has nothing to do with words. Text can be rendered but when writing equivalents, you need to write words. Using MathML is text and other stuff. Using just English to describe the picture is text and words.

/* Annuska Perkins and Gian Sampson Wild join by phone */

/* SBP finds


GV /* goes back to a higher level discussion...we got into details. */

GV Structure separate from vs structure provided. Split into 2. Everyone should perceive structure. Adding structure added to navigability/readability.

GV Operability, ensure that interface elements are operable by any user. Provide keyboard access. We had device independent handlers, it seems that keyboard access is what you get. /* discusses why created the grouping he did */

GV You can take a document, they can go through, provide descriptions, run it by the author to make sure the alternative forms are correct. Perhaps the author put it there not to be described but for some other fucntion. Perceivable/Operable - could be done after the fact. ...Orientation/Navigation/Comprehension...changing the content. thus involve the author more, can't do after the fact.

GV Success criteria: some of us are worried that we are coming up with a list of things that are "good to do." i.e. you have to do them in order to satisfy the checkpoint or sufficient and testable. I've been separating testable into 2 categories: machine testable or high human interrelator reliability. Several humans should come up with the same thing. Like KB's comment.

GV checkpoint 4.3 - Have a "conscious" checkpoint followed by a minimum requirement. The person has to consider the other things we thought of. But it's not a requirement. The first is allowing them to sign off on what they did. This is to help us get over the quandry of losing things we think is really important. We'll take heat since some people will say "the success criteria say do what you want" but we can say "do this at least, followed by do what you think is right."

LGR We have necessary but not sufficient criteria.

GV Right, as well as some that are not testable but important.

GV 1.1 - benefits...try to make it clear that this is more than just for people who are blind. /* reads the benefits */ Tried to bring things into the example sections that seemed to be floating around.

GV Thos issues that were not addressed, we left in. Others we got rid of. Others we added (from the issues list).

GV Digging through the archives and minutes was very time consuming. It is hard to find the original material since some people clipped out the original posts.

CS I ended up with a similar approach. "design for backwards compatibility"..."You have documented..." it gets at "decide what level you support and why." It's a similar idea of "think about it, do what's right, here's some info to guide you."

GV If we can say what the line is, specify it.

CS Baseline browsers, we can't say. It changes over time, is different for different localizations. /* reads some of the success criteria for one of the proposed 4.x checkpoints */

PB High level restructuring is positive. I posted a thread about a month ago, to reconsider the wording of the main guidelines.

GV yes, it was helpful.

PB Glad to see that some of my thoughts were incorporated. However, I had condensed them down into 3 guidelines. /* discusses differences */ Under orientation/navigation I had grouped 3.2 as a comprehension issue. I'd like to discuss that. Looking at the main outline.

GV Several of these we went back and forth. When we looked at the criteria, it said "emphasis structure" so people could "orient." Orientation is critical to comprehension. Here, trying to get at the thoughts as to why things are the way they ended up but not that they have to stay that way for that reason.

PB I can understand that. I was just curious. Additionally, the short word titles should be the same part of speech. perceivable vs comprehension.

GV Tried, but ...first are states others are facilitate. Achieve versus work towards. Nothing will be comprehensible to everyone.

JS You can't make it perceivable, you can't make me perceive it. only available.

GV Perceive is different than perceivable.

PB Nervous of losing "user needs and preferences" as a statement or category.

GV If you are telling teh author to do something,

PB Allow for user needs and preferences...

GV That's all 4 of them. Preferences is how deals w/rest of teh world. If don't need it not an accessibility issue. it's not to make things in the preferred form, it's in what they need. beyond that we have gone into usability. preference settings is usually what people need to use, therefore that term is confused.

PB In essence I guess that's what you mean. I can live w/out it, with that explanation. Configurable might be a better word.

GV If they can use it, then you don't have to make it configurable. Get into this more later in section 5.

KC When you are trying to do these types of classifications you are working out the responsibility of the author versus the deliver of goods and services. We get that confused. Legislation is always on the provider of goods and services.

CS an example?

KC Internet banking - legal obligation on the bank not on the person creating it.

GV More and more, the page will have services...e.g. banks have calculators and such that they didn't write, services

CS But the person who has to do the work is the author.

GV this is a large question to add to our list.

/* bob regan taking list of questions */

/* missed part of discussion */

MM Robust and evergreen are 2 words i'd like to use. Guideline 5. Robust, evergreen, future-proof. past-proof. durable.

ASW "perceived by any user" what if blind, deaf, and have mobility issues.

GV there are ways of doing it. as long as you have the info as text and the people can understand words if you talk with them. i've had clients who were blind, deaf, w/cerebral palsey, and mental retardation. can't do much there. cognitive...they can still perceive but not understand. we can get the info so everyone perceive.

CS can we say "with appropriate AT"

ASW Perceivable by one sense?

GV It does say that. Please read it.

MM 1.3 orientation/navigation

GV Content can be perceived across senses.

AP some of these checkpoints might fall within more than one guideline. maybe there is a way to provide that secondary mapping for people who are interested.

GV like a "see also"

JW a criticism of 1.0 is that there were implicit cross references that were not made in the text.

Flash demonstration by Bob Regan

/* Bob demonstrates a Flash application about Zoot culture working with Window Eyes */

$Date: 2002/03/29 21:05:38 $ Wendy Chisholm