Re: Revised Checkpoints: WCAG(1.4/1.3) and UAAG(2.5)

Jason White wrote:
> 
> I disagree completely with the approach which Robert Neff is advocating in
> this discussion. The priorities are to be determined on the basis of
> impact (as Eric calls it).

One moment, please. I think this is an important discussion.

1) The WCAG WG clearly established a scheme in which priority
   levels are based on impact.

2) The ATAG WG adopted that scheme except that some participants
   in the WG felt that other schemes would also be useful. If
   I recall (and I may not, so excuse me), one scheme might have been:
   "Priority 1 means that it must be satisfied to meet
   the needs of expert users; Priority 2 must be satisfied to
   meet the needs of average users; Priority 3 for novice users"

3) Robert Neff has suggested that implementation difficulty 
   should also be considered in deciding a checkpoint's priority.

I think it's very interesting to consider that other models
than "impact on the user" are conceivable. In fact, we should
embrace them to make our work easier and to keep our process
transparent as much as possible.

Consider the following ratings for checkpoint 1.1 of WCAG
1.0:  "Provide a text equivalent for every non-text element."

  Priority 1:     Means: Vital to removing barriers to accessibility

  Difficulty 3:   Means: It's easy to provide text equivalents that
                  aren't synchronized with other content
 
  Standard 1:     Means: Browsers and assistive technologies support
                  text today and support consistently.

Summarizing:
  - "Priority" refers to impact on users
  - "Difficulty" refers to effort required by authors or developers
  - "Standard" refers to current software support status

Comments:

1) There's no need to have a 1-2-3 scale for each rating model

2) There could be other rating models than these. For instance,
   the "content vs. user interface" division is useful in both
   ATAG and UAAG.

I think we all use more than one rating model for choosing
the priority of a checkpoint, though often the rating process
goes on silently because it diverges from the "impact" model
we currently use. By exposing the models we silently employ,
I'll bet:

1) We can avoid a number of discussions that in fact stem
   from the application of different and unspoken rating
   criteria;

2) We can provide more useful information to readers (though
   not necessarily all in one document);

3) We can systematize the way we think of checkpoints and
   add them to the guidelines. 

I do not mean to complicate the actual priority 
definitions; I think we should stick with "Priority" 
to mean "impact on the user". However, it's not only worth
considering alternative models, I think we would benefit
greatly from a fresh approach.

  - Ian

-- 
Ian Jacobs (jacobs@w3.org)   http://www.w3.org/People/Jacobs
Tel/Fax:                     +1 212 684-1814
Cell:                        +1 917 450-8783

Received on Thursday, 16 December 1999 21:21:09 UTC