Re: AERT Technique 13.6.1 and the WCA:UA agreements of yore

[Gregg and Jason:  you have been copied because by the end of this I say
that there is still a coordination issue and it involves coordinating with
GL.  Has to do with the tradeoff of using one element (i.e. MAP) vs. a
property (e.g. class="navaid") to capture and convey some role information
to user agents.  [Role definition could be made more clear, as well as
current and evolution-path means of incorporating that role information in
marked up documents.]

At 07:00 PM 2000-07-23 -0400, Leonard R. Kasday wrote:
>Al asked
>
> > When confronted with an arbitrarily deep nest (tree) of lists, do you mean
> > one should wrap
> >
> > 1) the innermost only
> > 2) the outermost only
>
>Ian replied:
>
>Yes, the outermost only is what I meant. That may
>not be sufficient for deep nested lists. But authors
>should probably avoid long lists of links anyway.
>

[and Len picked up the thread]

>Actually, it doesn't even work for one level of nesting... like I 
>mentioned, map inside map is an HTML 4.01 syntax error.

AG::

That's not what Ian said.  He said a single MAP containing arbitrarily many
nested list structures.

>I think there are reasonable uses of nested links, e.g. subject headings 
>and a few links under each heading, which is a very common idiom in 
>portals.  It's also arguably an accommodation for people with motor 
>disabilities since it reduces mouse clicks.   So I don't think we can 
>ignore this limitation of map.

AG::

Just a technicality:  You can't nest links.  You can't have another link
within the content of a link.  

However, I believe you mean a multilevel groupings of links.  This is a
common structure for a table of contents navbar.  Such a multilevel
grouping can be inside a MAP if I recall correctly.  

If you feel MAP does not accomplish some necessary job, this is something
to take up with the Web Content Accessibility working group.  But a clear
statement of the required job to be done and a realistic example of a
problem page where MAP falls seriously short should probably be offered to
support the suggestion to reopen this question.

Actually, I would like you to do that so there are two voices in opposition
to the comment that "there is nothing to coordinate."  Doing this right
ripples all the way to "what linking-to-schemas functionality is included
in XML Core?"  And we can't move ahead "around" the GL group; they have to
be involved at every step and on board if we move out with a position.

This is a tough issue to manage.  We need to correct the misconception that
CLASS values reserved in a PROFILE are anything strange or inappropriate,
so that more effective ways of indicating semantic structures which cross
threads with the syntactic structure can be duly marked.  But this is
subtle, it involves how the GL group is interpreting the formats which
raises PF issues, and we have to look at this as a migration from the
present state of affairs where the browsers don't support dictionaries for
CLASS to some sort of a future where layering data dictionaries over XML
grammars is a known and widely-implemented capability.


Al

>
>Len
>--
>Leonard R. Kasday, Ph.D.
>Institute on Disabilities/UAP, and
>Department of Electrical Engineering
>Temple University 423 Ritter Annex, Philadelphia, PA 19122
>
>kasday@acm.org
>http://astro.temple.edu/~kasday
>
>(215) 204-2247 (voice)  (800) 750-7428 (TTY)
>
>The WAVE web page accessibility evaluation assistant: 
>http://www.temple.edu/inst_disabilities/piat/wave/
>

Received on Monday, 24 July 2000 12:26:39 UTC