W3C home > Mailing lists > Public > w3c-wai-gl@w3.org > January to March 2003

RE: [techs] proposal to replace use of "rules"

From: Jason White <jasonw@ariel.ucs.unimelb.edu.au>
Date: Wed, 29 Jan 2003 14:57:16 +1100
Message-ID: <15927.20764.222572.268922@jdc.local>
To: Wendy A Chisholm <wendy@w3.org>
Cc: gv@trace.wisc.edu, w3c-wai-gl@w3.org

Wendy A Chisholm writes:
 > 
 > Thus, I'm wondering, "What if we moved the success criteria to the 
 > technology-specific documents?"  Here's an idea of what it might look like 
 > at each of the layers:
 > 
 > In the Guidelines/Checkpoints we might have:
 > Checkpoint 1.3 Make all content and structure available independently of 
 > presentation
 > 
 > [link to] technology-specific success criteria

This won't work. We would in effect be moving the normative success
criteria into a non-normative document. We would also be implying that
general (non-technology-specific) success criteria for this checkpoint
can't be written. We would, furthermore be leaving it to techniques document
authors, including potentially third parties in the case of
technologies for which the WCAG working group doesn't write
techniques, to define the success criteria for this checkpoint. Not
only does this give rise to the problem of non-normative documents
surreptitiously imposing normative requirements, it also gives
enormous discretion to the developers of the techniques documents in
defining what the checkpoint means. Finally, we would thereby depart
from the principle of providing technology-independent success
criteria for every checkpoint, a basic requirement underlying the
whole WCAG 2.0 project.

I propose instead that we (1) write technology-neutral success
criteria for checkpoint 1.3; and (2) include a non-normative note
referring to the Techniques for the details of how the criteria apply
to specific technologies essentially of the kind that Wendy suggested.
That is, I support the proposal to include a reference to Techniques,
but object to the idea of abandoning the construction of generic
success criteria in respect of the checkpoint. I think the current
level 1 success criteria capture important aspects of the checkpoint.
We could either just add a note referring to techniques, or add a
success criterion requiring that structural elements (elements which
convey semantics) are used correctly and occur wherever the
corresponding structural distinctions would be apparent from the
presentation. This would tie 1.3 more closely to the correct use of
structural elements, a connection which some reviewers rightly pointed
out was not explicit in the success criteria as currently written.

I also think we should leave considerations of simplicity and
comprehension aside for the moment and just concentrate on defining
what the technical requirement is, after which we can fine-tune the
wording to create as clear a statement as possible.
Received on Tuesday, 28 January 2003 22:57:34 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Monday, 7 December 2009 10:47:21 GMT