RE: User agent support of SUMMARY attribute in tables

> -----Original Message-----
> From: David Poehlman [mailto:poehlman1@comcast.net]
> Sent: Wednesday, January 22, 2003 8:49 AM
> 
> perhaps we can persuade the folks on bobby development to add 
> some table
> analesis that does not require that the summary attrib be 
> present in an lay
> out table. 

As some posts to this thread have mentioned, the WCAG appears to state that
all tables in HTML, whether used for data or layout, should have the summary
attribute. We have attempted to align Bobby to the guidelines as much as
possible, and careful consideration of the guidelines led us to incorporate
this interpretation into Bobby's evaluations. It has not been clear to us
whether conformance to that interpretation leads to the most useful result,
and from this discussion it is obviously quite difficult to decide this. But
conforming Bobby to the guidelines, to the best of our understanding, is the
best way to keep Bobby "honest".

Sometimes alternate interpretations or new information have been brought to
us and we have changed the support for Bobby, leading to some important
improvements. In the case of the summary attribute for tables this is a
thorny enough issue that I don't feel we can yet act on it. But the
information and ideas that have been raised in this discussion are very
useful to the process of evaluation development.

As Charles mentioned, the Techniques is an appropriate place for approaches
to issues like this to be represented. I want to second his suggestion and
encourage people to contribute their thoughts on this for formal discussion.
At Watchfire we also desire clear  understanding of the guidelines, because
as much as authors need to know how to structure content we need to know how
to evaluate it, and we intend to be a major contributor to the development
of Techniques for WCAG 2.0. We will design future versions of Bobby to
support those documents and it is important that people knowledgeable in the
field contribute to those documents to make sure their suggestions are
complete and realistic.

I would also like to take the opportunity to comment briefly on the
rationale behind the "Bobby Approved" statement in the report, which for
some is misleading. That approach to reports was originally introduced not
to target Bobby to managers, but to provide a "pat on the back", an
encouragement for work well done. It is contingent on passing certain
evaluations, with the knowledge that others still needed to be performed
manually to achieve full guideline conformance. This is a subtlety in the
reports that is missed by many users, and we continue to try to improve the
clarity of this. Our expectation is actually that Bobby would be of primary
use to individual developers, though Watchfire's Enterprise version will
also contain reports, new to Bobby, that are specifically oriented at
managers.

Michael

Michael Cooper
Accessibility Project Manager
Watchfire
1 Hines Rd
Kanata, ON  K2K 3C7
Canada
+1 613 599 3888 x4019
http://bobby.watchfire.com/

Received on Sunday, 26 January 2003 20:19:38 UTC