Minutes from 27 November 2000 telecon

Thanks to Chris for taking these.

Minutes available at: http://www.w3.org/WAI/ER/2000/11/27-minutes.html

27 November 2000 ERT WG Telecon
Participants
·       Brian Matheny
·       Harvey Bingham
·       Sean Palmer
·       Len Kasday
·       Wendy Chisholm
·       Chris Ridpath
·       William Loughborough
·       Dick Brown

Topic: Face To Face Meeting on December 4-5
LK: This will be a joint meeting with PF group. First day is ER only, 
second day is with PF. (Len can only attend first day.) We will also be 
coordinating with AU. Anybody know how busy is PF?
WL: Lots of overlap between us and PF. PF has plenty to do because of W3C work.
LK: Ideas for day 1 topics?
WC: Recent email on list, titled "Requirements for ADL", had questions 
about ADL. Perhaps we could discuss these questions.
LK: Not all questions answered.
WC: Topic could be to answer questions.
LK: Yes.

Topic: Accessibility Description Language (ADL)
WL: ADL - Why was it created could be topic. Design of ADL should be addressed.
LK: We should discuss high level of ADL and implementation details.
WL: High level is clear.
LK: What should it apply to? HTML and XHTML? Any XML, like SVG, it should 
apply to.
HB: Lots of XML is database to database and should not apply to that. 
Should only apply if doc has end user.
SP: All user interface technologies - it should apply to.
HB: If eventually sees user then it should apply to.
DB: Schema must have setup for access info.
SP: If user interface is not needed then accessibility info is not needed.
HB: Yes, but may only know at source database. If database is serving 
images then should have ALT text.
SP: yes.
LK: Yes, even database to database should include ALT text.
LK: What sort of guidelines are needed for ADL?
WC: Create semantically rich documents. Like allow images with ALT. 
Language should support guidelines.
LK: ADL general framework - statements point to something like elements, 
attributes or whole site.
WC: Two different things - schemas and content. WCAG 2 has new checkpoint 
that says "Conform to XML accessibility guidelines", so this is needed.
WL: XML Accessibility Guidelines!? (XML AG)
WC: PF is doing it.
LK: Is publicly available?
WC: Yes, link but is hidden from ER page.
WL: Is this DD note?
WC: Yes. But is PF work so they should do it. We don't know outcome. Will 
likely come back to WCAG but don't know for sure.
LK: Should ADL work on XML? Yes. Anyone see a complication?
Everyone: No.
LK: What about XML with syntax errors?
SP: Should not be syntax errors if conforms to WCAG P1 guideline.
LK: Even if has syntax errors we should be able to point to errors and note 
it is bad.
WC: We must leave in syntax problems - people may put them there for a reason.
LK: Even CSS. If well formed XML then can use Xpath. CSS, ECMA script, 
HTML, or XML with syntax errors.
WC: XML if not well formed then browser will crash. Tool could be same way. 
Or could clean up.
LK: Or tool could say "sorry invalid. Try again later".
WC: Question is "will we address" and I think answer is yes.
LK: If we see huge problem then do we address?
WL: What is ADL then?
LK: Specific language spec.
WL: Output of Wave, Bobby Aprompt?
HB: Static list or environment where repair can be made.
WC: Implementation detail.
HB: Standard way of reporting (can be shared among tools)?
WC: Yes.
LK: scope - defining language yes. Should also include presentation details?
WC: Could be scope of group in implementation, but not in language.
LK: Group will have at least one implementation.
WC: What is this going to be? A note or W3C recommendation? Must have at 
least 2 implementations for recommendation.
LK: For example, AU can't be recommendation candidate?
WC: AU went through before 2 implementation requirement was in place. Used 
many different tools to show that they can be done. Candidate requirement 
is now stronger.
WL: How did RDF get through?
WC: Don't know.
SP: Schema has been in for about a year.
WL: RDF is dependent on schemas getting accepted.
LK: Why need schemas for RDF?
WC: Schema is like table definition. When apps communicate, schema will lay 
foundation for discussion. Like DTD for HTML - "here's what you can expect".
DB: If no place for ALT then ALT will not be there.
WL: Web is like database so schemas could define data.
WC: Right. Like early days of HTML when browsers made best guesses. Can use 
RDF without schema but schema is good for strict typing and namespaces. 
More audit trail and can avoid anarchy.
LK: Essential for RDF like other XML languages.
SP: RDF schemas are not XML schemas.
SP: Just proposed specs for RDF. Maybe writing for Dublin Core.
WL: Are we like Dublin core for accessibility?
WC: Charles has proposed something similar. Expresses if complies to WCAG 
but also could include meta data about site.
WC: Can attach attributes to anything that has a URI.
LK: More than just attributes? Is ADL more than just attributes?
WC: What kind of inferences or logic would you like to include?
LK: Examples: what expert says etc.
WC: Like a database. Could perform a comparison of statements and facts.
LK: ADL could be a simple XML app. Or could be RDF app. Not implementation 
detail.
WC: XML tables of data. RDF expresses relationship between data. So is not 
either/or but likely both. (Both are XML.)
LK: Could be nothing but RDF.
WC: Right. We're looking at scope now and need to know what to bring to PF.
SP: Anybody seen Tim Berners-Lee 'design issues' document?
Everyone: No or very little.
SP: Will send URL to list. Tim & Dan describing web problems. Describing 
semantic web is very well laid out. Applies to what we want ADL to do. 
Parallels to ADL.
LK: Can anyone think of an example of accessibility description that avoids 
RDF?
WC: A-Prompt does it?
CR: New Java version does it (uses XML only).
LK: So what does RDF buy us?
WL: The 'f' (framework).
SP: XML is just data. RDF is metadata. Since ADL is meta data, why create 
more XML? RDF gives meaning so that's why it's needed.
WC: Interesting question to PF. Since PF reviews all W3C specs they see all 
problems and they can think about this.
WC: PF can check our thinking and see how it fits in with other work.
LK: HTML already has Meta tags.
WL: Backwards compatibility, we want forwards compatibility.
LK: What does RDF buy us?
SP: RDF is more universal (will work in not HTML files). Meta tags does not 
mean anything unless you know the XML but schemas are more universal.
LK: Self documenting, applies to other that HTML.
WL: We get more aggressive about our data about data.
LK: HTML spec has info about how to create doc that doesn't validate.
SP: Modularization is needed for other namespacing.
LK: Anybody working on module to put RDF into XHTML?
SP: Yes. Needed by others.
WC: Put on agenda - updates on XHTML and other info about this.
WL: Do have attribute in Head in XHTML for putting in metadata. What about 
the 'profile' file?
SP: Works for now. Can put RDF in to XHTML and is now used in W3C home page.
WL: What use can be made of it?
SP: Rich summary of W3C page. Can be used for extracting info about 
accessibility. Example - could be used to find W3C WCAG compliance logo and 
create WCAG compliance statement.
LK: If page is XHTML then could get tools to work on accessibility checking?
SP: Yes.
LK: (back to agenda) We need talks on technology.
WC: Yes.
SP: May attend by phone. What is needed by me?
WC: Example of what RDF buys us.
LK: Please put URIs on mailing list. Should we brainstorm ideas on what ADF 
will be like?
WC: Useful and instructive.
LK: Future discussion on list.
WC: Ask people to provide test pages.
LK: Break into 2 or 3 groups and compare results. Will continue discussion 
on list.


--
wendy a chisholm
world wide web consortium
web accessibility initiative
madison, wi usa
tel: +1 608 663 6346
/--