2 May 2002 - WCAG WG Teleconference




People vs individual/persons

Resolved: leave until text is more stable.

WCAG 1 to WCAG 2

Resolved: resolve on the list...hard to read the 3 paragraphs over the phone. Please indicate on the list by next Wednesday (8 May 2002) which of the 3 alternatives you prefer.

Conformance in general

WAC reads the conformance statement.

JS As WAC read, I find it troublesome that minimum conformance involves "removing barriers" which sounds like legacy sites rather than new sites. In new sites, would not remove but design to avoid barrier. Beyond minimum, there are some checkpoints where we don't have strategies for going beyond the minimum.

CS The text should reflect that one criteria for selecting the minima is based on testability.

Resolved: people need time to review the document. please take until next wednesday (8 May 2002) and submit issues for next thursday's discussion.

Action JS: write a sentence for the conformance paragraph that reflects the decision that minimum criteria are testable.

Parts of conformance statement

JS scope refers to which parts are included in the claim? circular defn.

JW This from WCAG 1.0.

WAC Defn of site is the next issue.

Resolved: seems reasonable to add these items. They are similar to the WCAG 1.0 conformance statement.

/* GV joins */

Definition of "site"

Suggested definition

For the purposes of these guidelines, a site is that body of web content which the conformance claim encompasses.

GV My concern is that someone could put a sticker on a home page that says they conform. But, the claim is only for the home page not the whole site. Probably effective.

CS I agree that there are situations where we don't want to define by domain. e.g. /~ where one site is conformant and the rest are not. The other, big sites that spam multiple domains or servers.

LR A body of elements in a domain claiming conformance to WCAG 2. It doesn't matter if it is a college where one area wants to make a claim and the others don't. They are still in the same domain.

CS It can spam multiple domains.

LR Not everyone is that large.

CS Don't constrain it to domain.

JS GV's defn would cover that.

/* agreement */

ASW How would you identify your site?

JW In the conformance claim, you define it.

CS Say something like "i'm claiming conformance for everything on foo.com, or everything msn.com and passport.com, or everything under foo.com/~me, etc."

JS Our human resources is reasonably conformant, but draws on services from groups who are not. Thus, they claim conformance for pages they own, but hard to tell from user's point of view.

CS Done by set of URLs?

JW Anything maintained by person X or department Y.

GV What if that section of the page is not maintained by.

CS Needs to be a mechanism for explicitly excluding or including a page. Ownership is not a good rule to expose to the user. It's like a privacy policy.

CO We (in St. of Virginia) have to be clear about when people leave our site.

GV If we have a site that is accessible, but each time to place an order it jumps off to another site that processes the orders. Then your site can be accessible, until you get to a service organization. That would be hard for that site to claim conformance (summary: a site has a way to browser products and claims it is accessible, but once you want to buy you go to a site that is not accessible. therefore, you can't really use the site in a meaningful way unless you only want to browser products).

JW Common sense needs to be applied.

WAC Perhaps another checkpoint (or guidance somewhere in the guidelines): if your site combines content from a variety of sites that combination ought to be accessible.

JS We had that problem here. Wrapper was accessible but the core content were applets that were not accessible.

JW If people make a claim then they decide for what they are making the claim.

CS And we give them the tools to make it accurately.

MM If you incorporate content from another source, then that content needs to be accessible. That should cover syndicated content that is not being linked to but the core of what the user is going for.

GV sounds great. Why can't a generated site meet all of the guidelines.

CS Perhaps you wouldn't meet all of the cognitive guidelines or rendering issues...

GV But you could meet the cognitive, just not while meeting the others?

CS On a small screen scenario, you can't do keyboard access if no keyboard.

GV All small devices have a keyboard of some type.

CS I can't think of a good example right now.

Resolved 23 May - we'll discuss generated content and different version of content. It would be good to make sure that Kynn, Scott Luebking, etc. are present. WAC takes the action to remind people to prep for the upcoming discussion.

5 Magic words

GV last one be durable?

CS Better than anything else we've come up with.

JS What about "adaptable?" I think we're talking about pages that designed to work with whatever technology people throw at them.

GV It means "able to be adapted."

JS Durable means persistence over time.

CS Flexible?

MM Tried that, was a bit too nebulous.

JS I'm ok w/robust but it doesn't end in "-able."

GV They don't have to have "-able."

MM "Robustable."

/* laughter */

GV Perceivable or perceptible?

MM Perceivable more widely understood.

JS As long as it is in the dictionary i'm fine.

Keyboard input vs character input

Checkpoint 2.1 [2.5] Provide keyboard access to all functionality of the content.

CS Tab key is character sequence.

JW This does not cover voice input systems and device independence. If you said keyboard-operable that would mean if you used keyboard events that would work, but many devices might not be able to emulate that.

GV anything done w/speech also has to be done through keyboard, since people who do not speak need keyboard input. "Encoding input" is another option, but it is cryptic.

CS Palm support a plug-in keyboard?

GV Yes.

ASW Can you navigate or only input text?

GV I don't know.

ASW I've heard that you can't

MM There are graffiti strokes for movement, technically in graffiti means to navigate.

GV What about an external keyboard?

JW Thinking about VoiceXML work.

WAC finds VoiceXML appendix H - Accessibility that says VoiceXML only provides aural.

JW There is a difficulty of reliance on things that might not be device independent.

CS How about, "for content that is intended to be rendered on a device w/a keyboard, provide keyboard access."

GV For others only do pointing.

CS We're back at multiple forms. If you need a keyboard you should use a device that has a keyboard. If you are building a web page that you expect to use in a PC browser use shortcuts, in WAP don't use.

GV Why doesn't WAP version need to have a keyboard? All phones have keyboards.

CS If using a voice browser, do you assume this is the way you use apps w/a keyboard?

GV What if we said operable from alpha-numeric + tab/shift/enter/backspace functions?

JW We get into internationalization issues.

GV I said alphanumeric I didn't say whose.

CS Any character that can be encoded in unicode? ... what about event model?

JW What is this trying to capture?

GV Evolved from device indie handlers to keyboard. why?

CS Making it clearer and simpler.

GV Are device-indie events HTML specific?

WAC No. XML apps (like SVG) are also defining events.

/* discussion about events. */

GV Operable through text entry?

JW Abstract device-indie event.

CS 2 aspects of keyboard operable - put in text or move by keyboard. e.g. onfocus can happen with an event whehter from mouse or keyboard. Could put your name into the form through speech (once speech recognition is up to snuff).

GV We talked about device independence. It is a misnomer. For text-entry, can't have device indie, you can't use a mouse you have to have something that generates text.

CS One case that isn't feasible now but will be soon is voice recognition.

GV Isn't that text entry?

CS Whether it turns into text in an intermediate stage is an implementation detail. It wouldn't necessarily have to. It could be pure speech. If that were the only way to put data into that form it would not be accessible.

GV Therefore, everything should be operable through text-entry or do we have a definition of device-indie.

JW Unicode+certain operations must be available.

GV Does everything in device-indie boil down to eye-hand-coordination.

CS Right now, I think it is, but as speech systems get better, there will be systems that a person who is deaf could not use.

Action WAC: I'll be on a device indie panel next week, I can take this past them.

BF reads from Device Independent Principles

For some Web content or application to be device independent, it should be possible for a user to obtain a functional presentation associated with its Web page identifier via any access mechanism.

GV Does that mean as long as there is a mouse, it is ok? Keyboard or mouse means could only do with a mouse. As you get to the root, the terms and defns are used euphemistically. When we say device indie, what do we really mean? It doesn't mean that you could operate w/any device. Don't really mean mouse alone.

$Date: 2002/05/02 22:45:04 $ Wendy Chisholm