W3C home > Mailing lists > Public > w3c-wai-gl@w3.org > April to June 2001

Re: 4/26 WCAG meeting minutes

From: Anne Pemberton <apembert@erols.com>
Date: Thu, 26 Apr 2001 21:35:06 -0400
Message-Id: <>
To: w3c-wai-gl@w3.org
Cc: kshea@apollo.fedworld.gov
A thought on separating normative guidelines from usability guidelines ...
Will there be a normative guideline to call for an ALT tag, and a usability
guideline to say that the alt tag should "describe the function of the
graphic" ?????


At 03:33 PM 4/26/01 -0700, Loretta Guarino Reid wrote:
>In attendance:
>Loretta Guarino Reid
>Andi Snow-Weaver
>William Loughborough
>Greg Vanderheiden
>Cynthia Shelly
>Jason White
>Paul Bohman
>Gregory Rosmaita
>Katie Haritos-Shea
>Charles McCathieNevile
>Action Item:
>GV,JW: Divide guidelines based on Greg's proposal to divide normative 
>accessibility guidelines from recommended usability guidelines.
>JW: Checkpoint 2.1 (provide more than one path or mechanism to find content) 
>is still listed as an open issue. How do we decide under what circumstances 
>it applies?
>CS: Is there a doubt about it?
>JW: It doesn't seem to apply universally. There is some content that isn't 
>so long or complex that it needs links, search mechanism, etc. Is there a 
>complexity threshhold for this? What else can we say about it? How can it 
>be stated so it is clear it doesn't apply to every page. 
>The second issue is whether the checkpoint can be defined more exactly.
>GV: I'm uncomfortable with this.
>JW: So am I, but probably for different reasons.
>GV: My problem is that we started off with guidelines for accessibility, and 
>now we are getting deep into usability, and I think we have no business
>And for each item we add, we need to realize we are weakening every other 
>item. I will be leading a movement to try to elimate about a third of the 
>guidelines. I feel we've lost focus on accessibility issues. Or we need to 
>divide this into two parts, one section for accessibility and one section 
>for web design, good advice, etc. Our definition of priority 1 is that
>the guideline, there are people who won't be able to use the content. If 
>we don't make pages targeted for people with an IQ of 40, they won't be 
>able to use them. If we don't make these guidelines priority 1 for these 
>people, they'll be excluded. Anytime a website isn't dead simple, there 
>will be some people who won't be able to use it, but we can't make all 
>websites dead simple. We are adding more and more checkpoints addressing 
>usability issues that are only accessibility issues for people with 
>cognitive disability. I worry about how providing more than one path makes 
>information more accessible. It makes it more usable.
>JW: My problem is that this guideline looks, not technology specific, but
>specific to a certain way that web content can be organized. The examples 
>seem strongly oriented towards the mixture of UI and info that you get with 
>contemporary HTML and don't know if this is what the web will be like
>over the next several years. 
>CS: Are we being too specific with these examples? because over time these 
>solutions will change as technology changes
>JW: These examples are things that characterize HTML very well but may not 
>charcterize other technologies very well.
>CS: Some examples seem like they apply more generally.
>GV: "Jump over a link" stands out as having been stuffed in here because 
>it needed a home.
>JW: It goes under techniques
>GV: It used to go under group items.
>JW: The HTML checkpoint solutions is where it would live; it is also going 
>to become obsolete.
>CS: But that is not a reason not to address it. It won't be become obsolete 
>for a while.
>JW: But it still belongs under HTML techniques.
>WA: What are we addressing with this guideline?
>CS: This is mostly a cognitive issue, to make it easier for people to 
>get where they need to get as easily as possible. I understand what Greg is 
>saying about usability, but there is a lot of crossover. It is advantageous 
>to get some usability in.
>WA: I'm not sure this gets us usability. I don't understand.
>??: If you have a big document, you need a summary and some way to get the 
>information other than just reading a page at a time. You should provide 
>a table of contents or index or search function. 
>CS: There is debate in the accessibility community about whether it is 
>better to provide lots of navigation functions or one simple option that 
>is easy to use.
>GV: You ought to be able to do a boolean search on a book. But you put 
>that functionality under an Advanced button, not right up front.
>CS: It has to do with organizing information. But I can't find the
>GV: There is one in WCAG1.0
>CS: What about the checkpoint about the logical structure of content. Maybe 
>that's where it should go.
>JW: Things should be marked in pieces accordingly. The User Agent should 
>permit you to manipulate the chunks for navigation.
>CS: Guideline 2 says provide interaction to suit users needs and preferences.
>GV: This should be a priority 2, and it should permit the user to move 
>efficiently between different parts of document. Guideline 2.1 should be 
>"provide mechanism to permit user to move efficiently through the content".
>JW: We are dealing with several issues: 1) the broader question of where 
>this guideline should go, and whether it falls into the class of 
>accessibility-related requirements, 2) what should the complexity threshhold 
>be for this requirement, and 3) is it sufficiently clear what it means? 
>The examples help clarify.
>CS: This checkpoint is understandable, achievable, and measureable, more than
>a lot of other checkpoints. This checkpoint doesn't seem particularly vague.
>JW: What about issue of how complex the content needs to be. The checkpoint
> appears to apply to everything.
>CS: It applies to site, not to a page.
>JW: if your site is 2 pages, does it apply?
>CS: I know about two rules of thumb used by real world web designers. Nothing
>should be more than 3 clicks from your home page. And no group should have 
>more than 10 items. It is usually impossible to satisfy both of these for a
>complex site.
>GV: How to you group states with these rules of thumb?
>CS: alphabetically? geographically? 
>GR: Most sites organize states geographically.
>CS: or in a list, in alphabetical order.
>JW: A long list
>CS: in English
>GV: If someone arranged the list of countries by continent, and even the
>UN wil have trouble finding some countries.
>CS: Almost no one with a real site achieves those rules of thumb.
>JW: How do we define the complexity at which this checkpoint comes into
>CS: 10 pages or more? 5 pages or more? It will be hard to define a hard rule.
>But I know it when I see it. This decision must be based on professional 
>GR: In the User Agent guidelines, we had a similar issue with the time
>on refreshes and pauses. If you choose any number, it will be arbitrary.
>It becomes an area for scenarios rather than requirements. 
>CS: Help people to exercise professional judgement.
>GR: On optional parts, you can only give guidance.
>JW: I think we have a proposal that for this guideline, we help people
>judgment by giving examples in the techniques documents.
>CS: and links to usability documentation. There are web sites and books on 
>these topics. Providing info is useful
>JW: Examples of application scenarios should be provided. "This applies 
>to complex content only". Example will help to develop reasonable judgement. 
>This solves the threshhold problem for the moment. I think of this as one 
>among a number of cehckpoints that individually may not make a difference 
>between accessibility and inaccessibility, but jointly they are likely to. 
>GR?: A comment on demarcation. The User Agent Guidelines has content types. 
>A checkpoint is marked with which content types it applies to. Don't divide 
>the checkpoints into 2 classes (object, subjective). Rather than grouping 
>checkpoints, label checkpoints.
>GV: Flagged with attributes, not sorted into categories. This is important 
>since we already have 3 attributes.
>CS: And there are person-to-person variations in ease to use.
>JW: I find that anything that requires reading through lists of menu items 
>or links is useful initially but becomes a nuisance thereafter. I tend to 
>use search. 
>CM: This is a feature of graphical browsers, e.g., ICAD, Opera.
>JW: What dimensions will we use for labels? A set that is supposed to make 
>it cognitively easier to work with content. 
>GV: I made an initial cut at labeling the guidelines as priority 1, 2, and 3.
>A couple are hard to decide. I came up with 9 priority 1, 5 priority 2,
and 7 
>priority 3 of 21 items.
>JW: Please send your categorization to the list.
>GV: It will provoke interesting discussion.
>GV: First we need to address when  usability is accessibility. For people 
>on the edge, any usability problem pushes them over the edge. This means
>that every single usability issue becomes an accessibility issue. But we 
>really can't go there. For example, for a ramp, 1 in 12 is defined to be 
>accessible, but there are lots of people who can't push up that levelof 
>GR: Braille labels on elevators can't be used by many (most) blind. And they
>introduce cognitive problems, too - which button goes with which label?
>GV: These are the nuances we don't want to lose. This is where we see the 
>difference between cognition problems and vision problems. We can't deal 
>with complete cognitive disability. It becomes a questino of how much of 
>the bottom we are going to chop off.
>KHS: We need to decide that question.
>GV: There will be a lot of flack when we do, but we need to do it. We will 
>exclude some people. I don't know where it starts or stops, though. If we 
>say we will only worry about people at a certain level of cognitive ability,
>we are doing the same things as the marketing people who say that are only 
>targeting customers without disabilities.
>KHS: We don't have to throw out the baby with the bathwater.
>GV: Consider amazon.com. How will we make this site accessible to someone 
>with IQ of 60?
>KH: We need to define minimal required cognitive level.
>GV: Will we require Amazon to make an alternative web site for those below 
>that level?
>WA: And a site with a target IQ of 75 is useless to someone with an IQ of
>JW: I agree with Greg that that's the problem. The techniques for cognitive 
>disabilities should be written out. But I'm not sure how to deal with them 
>in this document. Should they be prioritized specially? tagged someway?
>GR: If there is an application of WCAG to site in a legal juridiction, if
>site is challenged for not meeting cognitive needs, the courts will defer to 
>whatever is on the statute books.
>GV: My problem is that I want to develop a site for teachng thermodynamics. 
>I can't create a version for the  cognitively disabled.
>GR: There will be different threshholds applied for government entities, etc.
>GV: A university course has prerequisites. These should also apply to the
>GR: We need to collect these factors.
>GV: I suggest that we split the guidelines into 2 categories. The first 
>category is those things you must do to comply (at priorities 1,2,3) that
>are expected of all sites. Another category contains guidelines that you 
>should apply to the extent that you can, and especially if you want to 
>target sites for a particular population. This proposal gets us around 
>two points. A site can be fully compliant, down to level 3, without 
>worrying about this last category. But we don't abandon the cognitively 
>disabled. And we can identify who will be most aided by the second class 
>of items.
>JW: This could be useful. We could try to apply this to guidelines.
>GV: This last section would not be normative. They would just be 
>recommendations that you do if you can.
>WA: Will Greg divide the checkpoints into these two classes?
>JW: I'll help
>GV: We have to figure out how to provide guidelines that can be applied 
>to all websites. But we also don't want to remove everything that can't 
>be applied everywhere. e.g. the Sesame Street site shouldn't require the
> same cognitive level as an MIT physics course.
>JW: This is a better characterization than objective/subjective. Let's
>take an action to look at that division, see what comes out, and discuss it
>at a future meeting.
>JW: Let's discuss the face-2-face in June. We need to work out the details
>the next week or so. Jason and Greg aren't available, except maybe by 
>CM: I'd be happy to chair the meeting if Jason and Greg can join by
>As ATAG chair, I would like the HTML techniques to make progress. 
>GR: I want to make sure there is critical mass, unlike the AU meeting in
>JW: Wendy said she would be able to run this, if we go ahead with it.
>JW: Any problem with no co-chairs present?
>GR: Too good an opportunity to pass up
>KHS: Please review the PDF techniques spec. We need feedback on what 
>techniques go with which guidelines.
>PB: I have a question about Greg's proposal: are we reducing the
priorities to
>just priority 1 and 2?
>GV: No, we would still have 3 priorities. There would also be another 
>category of items that are usability guidelines that are recommended but 
>not normative. (Advice) 
>JW: When there is a question as too how much to do something, as opposed to 
>whether it was done or not, that guideline probably goes into the second
Anne Pemberton

Received on Thursday, 26 April 2001 21:47:48 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Monday, 7 December 2009 10:47:10 GMT