W3C home > Mailing lists > Public > w3c-wai-gl@w3.org > October to December 2002

RE: Review of WCAG 2.0 WD: structure found confusing, hiding impr ovements in simplicity

From: Jukka Korpela <jukka.korpela@tieke.fi>
Date: Wed, 9 Oct 2002 15:29:45 +0300
Message-ID: <621574AE86FAD3118D1D0000E22138A95BDECA@TIEKE1>
To: "'w3c-wai-gl@w3.org'" <w3c-wai-gl@w3.org>

I'll try to comment on Wendy's remarks on my message. Sorry for the delay in
doing so.

Wendy wrote:

> Could you please list which WCAG 1.0 principles are "homeless?"

I suppose I was mainly thinking about guidelines 10 and 11
("Use interim solutions" and "Use W3C technologies and guidelines"). They
are technically oriented and don't quite fit into the approach of the WCAG
2.0 draft. To put it in another way, guideline 5 in the draft now contains a
mixture of technology-oriented principles that don't quite play in the same
field as the other guidelines.

Moreover, WCAG 1.0 guideline 5 (Clarify natural language usage) has been
embedded into checkpoint 1.5 (Provide information needed for unambiguous
decoding of the characters and words in the content) in the draft. This
doesn't look very natural to me. The need for indicating the language used
goes far beyond decoding issues (and the word "decoding" is even partly
misleading here since it suggests connection with character encoding

> Could you give your reasons for wanting to move 4.2 under 1.1?

I'd say that 1.1 tells us to provide textual alternatives to non-text
content, and 4.2 is a dual principle of providing non-text alternatives to
textual content. There's a difference of course: the first one is
obligatory, the second one applies "where appropriate". This reflects how we
regard text somehow as the  solid basis of accessibility, as primarily
accessible. We probably need to think about this, deeper than we've used to.
Even if you don't make the two principles adjacent in the order, it was a
thought-giving experiment to consider them that way. (On the practical side,
I'm sure there are situations where textual content _needs_ non-text
alternatives or augments, in a sense comparable to how images _need_ textual
alteratives. For example, a merely verbal description of a bird species
might be crystal clear to an experienced ornithologist, but most of use
would really need some picture(s) and/or sound samples to understand what
the bird is like. But I don't know whether it is possible to formulate
general requirements on augmenting text with non-text content to make it
accessible to all.)

This relates to the general question about the relationship between
guideline 1 (Perceivable) and guideline 4 (Understandable). Could non-text
alternatives be part of the former, rather than the latter? Admittedly, text
content is basically "perceivable" in the sense that it can be presented
_somehow_ to a human being who is able to communicate at all. I think I see
the reasoning behind the division into Perceivable (to senses) and
Understandable (to the mind), but it wasn't very obvious. And it might have
been easier if Understandable were presented right after Perceivable.

> The technology-specifics for WCAG 1.0 are not normative, nor will the 
> technology-specifics for WCAG 2.0 be normative.

Maybe they should be partly normative. Naturally they would be conditional:
_if_ the technology  used is such-and-such, then the guideline this-or-that
shall be implemented so-and-so. This would often make the guidelines clearer
(more accessible!), since often it's the mapping to a specific technology
that makes people understand the intended meaning of a guideline.

> I'm concerned about your statement that the European parliament considers
> techniques normative.

I didn't mean the techniques document. My reference to WCAG 1.0 was somewhat
vague; I meant the guidelines document itself. The wording in the resolution
of the European parlament is:
"K. whereas the world wide web consortium set up the web accessibility
initiative (WAI) and the latter has developed the web content accessibility
guidelines version 1.0 called "the guidelines" which are nowadays considered
to be the global standard for the designing of accessible websites; - -
- -
[the parliament]
31. Stresses the fact that, for websites to be accessible, it is essential
that they are double-A compliant, that priority 2 of the WAI guidelines must
be fully implemented;"

I'm afraid I've presented my point poorly. I'll try again: Currently, the
WCAG 1.0 guidelines document contains rather specific recommendations, more
specific than the content of the WCAG 2.0 draft. The WCAG 1.0 guidelines are
widely regarded as "standards" or "normative" in some sense. If WCAG 1.0
guidelines and associated documents will be replaced by WCAG 2.0 that is at
the current level of specificity, with some things currently in the WCAG 1.0
guidelines moved to techniques documents, then e.g. the European Union will
probably update their documents to refer to WCAG 2.0 as normative. This will
imply that the status of some important principles will drop.

> People who try to implement WCAG 1.0 are often unable to 
> determine if they have met the requirements.

Yes, especially since many of them are not objectively verifiable, still
less checkable by computer programs.

> Do you have a specific 
> suggestion for how we can provide clear guidance to authors?

I don't know whether this helps anything in that respect, but I regard this
as important for other reasons, including guidance to people who write
accessibility checking and repair software: Technical rules (which are
partly implementable as automated checks) can usually only identify items to
be checked, not actual compliance with accessibility guidelines. Example: We
can specify rather exactly how one can identify non-text elements on a page,
and in part we can determine mechanically whether there is something
presented in a position where a textual alternative is to appear. But this
will not answer the question whether it specifies an adequate textual
alternative to the non-text content. That's something that a human being
needs to decide. The general thing we can say about this is probably just
that the alternative should work in different situations, such as text-only
display, speech presentation, and Braille presentation (and these might have
somewhat conflicting requirements). Similarly, a guideline on writing as
simply as possible/feasible could be made more operational by listing some
ways to test the understandability, like asking people with comprehension
problems to read it and explain the content. The guideline on providing
methods to minimize error and provide graceful recovery could be made more
concrete by saying that pages, especially interactive pages, should be
tested by people are likely to make mistakes (or by someone who
intentionally makes mistakes). But, unavoidably, this would normally just
list _examples_ of testing methods, and mostly the testing would not give
any decisive results. (Organizations might, of course, make some specific
testing methods required, especially if they can decide that some particular
tests are essential due to intended audience and content.)

Jukka Korpela, senior adviser 
TIEKE Finnish Information Society Development Centre
Received on Wednesday, 9 October 2002 08:30:15 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Monday, 7 December 2009 10:47:20 GMT