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

Re: WCAG1.0 Checkpoint 10.5 - still valid?

From: Dave Shea <dave@mezzoblue.com>
Date: Wed, 10 Dec 2003 09:06:57 -0800
Message-ID: <3FD752B1.9060106@mezzoblue.com>
To: Kynn Bartlett <kynn@idyllmtn.com>
Cc: w3c-wai-gl@w3.org

>> Where is it stated that a page author isn't allowed to make the claim?
> 
> It's not.  But there is no mechanism for making that claim, and no
> guidelines by which page authors are "allowed" to specify that such
> a state has been reached.

Which, I guess, leaves it open to a judgement call by the page author. 
Or at least, there's nothing specifically forbidding this as it stands.

> This is why every single checkpoint with "until user agents" is
> fatally flawed.  Checkpoints were given with expiry conditions but
> with no information on how those conditions can be met.

Of course, dwelling on your single Croatian browser example, there's 
always the question of who determines the cut-off points. WCAG, I'd 
assume, is inclusive as possible - but if page authors are building 
English sites, to be used by English readers, a particular point dealing 
specifically with Croatian user agents isn't that relevant. If it's 
hard-wired into WCAG, then more work is being expended by the author 
than necessary.

> To some degree, it can be left up to the author -- but then, so can
> all checkpoints.  Reducing accessibility to a series of subjective
> decisions made by Web developers with incentive for self-approval,
> who likely don't have much knowledge of assistive technology or people
> with disabilities is a Real Bad Idea.

Good point. User checks are inevitable in cases where software can't 
comply, otherwise the guidelines would quickly devolve into a silly list 
of rules like "always use #000000 text on a #cccccc background to ensure 
contrast isn't blinding while legibility is maintained."

> I wouldn't say the bug is that it's left open, I'd say that the bug
> is that the "until user agents" clauses don't actually give Web
> developers adequate guidance, and requires them to be experts in order
> to know whether a checkpoint should be followed.

This is a realistic evaluation. By the same token, I'd presume the 
'until user agent' checkpoints were created in a conscious awareness 
that the points mentioned would eventually become irrelevant. Perhaps 
the wording is wrong, but any shifts in the future should allow for 
updating if the clauses become useless.

Planned obsolescence. It's what got us here, but it shouldn't be 
ignored. A second edition of WCAG1.0 would go a long way toward 
satisfying questions like mine.

> This is why some of us have been pressing for the WCAG working group
> to issue a WCAG 1.0 Second Edition, to fold in errata and bring the
> "until user agents" clauses up to date.  You have correctly identified
> what I, as a quasi-outsider, see as a serious problem:  WCAG 1.0 is
> currently unworkable in the details, and needs some tweaking to restore
> it to a form in which 2003/2004 Web developers can address accessibility
> until WCAG 2.0 appears.

Well now, we ended up in the same place. That's saying something. :)

d.
Received on Wednesday, 10 December 2003 12:09:41 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Monday, 7 December 2009 10:47:26 GMT