Summary of Technical Plenary 4 March 2003

Here is a very brief summary of the technical plenary last week. A lot of
neat ideas came up that will give us a lot to work through. I invite anyone
else who was present to mention things I've missed. I apologize that these
notes are somewhat rough. Michael

Wendy gave a high level overview of the XML Schema we're using and features
of XML Spy for editing. People who are actively editing documents for WCAG
can get a license to XML Spy that Altova has donated.

We discussed briefly whether the techniques should be normative. We think
not. Other test suites are not normative, and we don't want to imply that
our techniques represent the complete set of all possible techniques,
especially given technology change.

Test suites came up and became a major focus for the day. Ideally, user
agents, authoring tools, and evaluation tools would all support the same
test suite though there may be practical issues. We want to follow existing
practices and spent time discussion materials used in the creation of the
CSS test suites and resources from the QA Working Group.

W3C process requires that two implementations of a specification be created
while it is in the Candidate Recommendation phase in order for it to proceed
to full Recommendation. We discussed what an "implementation" of WCAG is.
This could include user agents and authoring tools, or evaluation tools, or
even simply sample Web sites and test files.

After the initial discussion we broke into topical groups to explore issues.
Later in the afternoon the groups each reported briefly on their activity.

 - The group working on SVG found that it was difficult to apply the
guidelines as they exist to a medium like SVG. WCAG focuses on text
alternatives for media but their goal was to make natively accessible
graphics. The proposed that a Level 1 success criterion for graphics might
be to provide a text alternative but a Level 2 success criterion would be to
provide a natively accessible graphic.

 - A group working further on test suites and QA process proposed that we
adopt a QA framework. Steps include a) update WCAG WG Charter doc for QA
commitments; b) Pick a QA contact/lead for WCAG WG; c) create a QA Process
Document; d) adopt a way to have QA discussions on the list, e.g., by
including "[test]" in the subject line of posts; e) create a WCAG/ATAG Test
Suite Documentation document.

 - A small HTML group explored techniques for tables. Because it would be
desirable to provide mutually exclusive techniques for data and layout
tables, the issue of technique granularity and relationships of techniques
came to the fore. Drawing from some of the work on checklists, it seemed
desirable that techniques be dependent on certain conditions, e.g., "this is
a layout table" and "the table lays out a navigation bar".

 - The checklists group presented two approaches to checklists. One is a
fairly long format that documents all the conditions relevant to a
particular success criterion, but was several screens for a single success
criterion. Another is interactive, in which questions are asked and used to
filter what checklist items are presented.

 - The group working on RDF and metadata explored how to use metadata to
provide supplemental content so non-compliant documents can pass the
guidelines. They annotated changes to wording of the guidelines that would
be useful and mentioned that Guideline 4 is difficult and Guideline 5 poses
major problems for this.

After the breakout sessions we joined the Education and Outreach Working
Group. The main focus of this meeting was to identify coordination points
for future work as the EO group produces materials that will need to be
synchronized with WCAG 2.0. The groups will follow up with coordination
plans and identify some joint work.

Next we were joined by the co-chair of the QA working group to introduce the
QA process (see to us, request our feedback on it,
and offer potential resources. Some areas of work in which we might want QA
assistance are a) review process and specifications; b) provide consulting;
c) document existing best practices; d) provide templates, tools, and test
harnesses; e) provide feedback on test guidelines.

Finally we were joined by the Internationalization Working Group. They are
also working on guidelines, though with less of an emphasis on conformance,
and we want to share knowledge and tools. They presented some of the work
they've been doing. The issue of coordination not just on process but on
content also came up - that there might be accessibility issues in
internationalization guidelines and internationalization issues in
accessibility guidelines. We hope to establish points of coordination with
this group.

Michael Cooper
Accessibility Project Manager
1 Hines Rd
Kanata, ON  K2K 3C7
+1 613 599 3888 x4019

Received on Monday, 10 March 2003 15:16:25 UTC