W3C home > Mailing lists > Public > w3c-wai-gl@w3.org > October to December 1998

Re: Request for quick feedback to WAI EOWG (fwd)

From: Alan J. Flavell <flavell@a5.ph.gla.ac.uk>
Date: Sat, 7 Nov 1998 18:02:21 +0000 (GMT)
To: WAI GL <w3c-wai-gl@w3.org>
cc: w3c-wai-eo@w3.org
Message-ID: <Pine.OSF.3.96.981107171141.19466B-100000@a5.ph.gla.ac.uk>

I think the card checklist was an excellent idea.  As another f'up
said, if _everyone_ did at least what it advised on the card, the web
would be a _great_ deal further forward. 

There's a widespread superstition around that making web documents
accessible means emasculating them for their originally-intended
audience. If there's one thing we simply have to get right, it's to
dispel this superstition.  Otherwise, it's my continual fear that the
only accessible web documents will be a small subset that are subject
to local accessibility legislation, or written by a few activists,
while everything else will be "designed for" one standard browsing
situation that's perceived to represent the "mass market", to the
deliberate exclusion of everyone else. I'm sure that the participants
on this list have every intention of "doing the right thing" with
their own documents, but it's all the other authors "out there"  that
worry me. 

Furthermore, with increasing use of authoring software (and I'm not
for a moment opposing that - it makes a great deal of sense), it's
important that this software actually inculcates good authoring
practices.  Software that does not automatically create valid HTML
syntax, and does not positively demand authors to provide ALT texts
for their IMGs and AREAs, and so forth, needs to be exposed (there is
clearly a lot of it about, judging from the web pages it produces -
I'm naming no names, readers can fill in the blanks from their own
experience, I'm sure).  So, the activity needs to target the designers
of authoring software, as well as authors themselves.

On Sat, 7 Nov 1998, Charles McCathieNevile wrote:

> I agree that validating HTML is more important than 
> checking in three browsers and using Bobby, since they are not foolproof 
> either, and it is not a scaleable solution.

I don't see this as an either/or.  It's possible for a misguided
author to produce syntactically valid HTML that is semantically
nonsense, as well as being highly inaccessible.  I support the use of
syntactically valid HTML, but it's only one aspect of authoring a web
page.  With experience and by following good practices, it's possible
to create web pages that don't have to be checked in three browsers on
three different platforms and fed to Bobby, but for those who haven't
yet garnered that experience (and I'm still learning, as I presume
everyone else is to a greater or lesser extent), I don't see any
alternative to these practical evaluations, at least now and again
on sample pages. 

> 1. Images need ALT to describe their function

Both on the original draft of the card, and in this proposal, the use
of the word "describe" worries me.  The ALT ("alternative") text is
supposed to be substitute text that can serve the same purpose as the
graphic that it is the "alternative" for, not to "describe" that
purpose.  Many of the ALT text "howlers" that I see on the WWW are the
result of someone who is providing a description of a function, rather
than the function itself. 

In general, I strongly support the statement in the HTML4.0 spec that
the purpose of the ALT attribute is to "specify alternate text to
serve as content when the element cannot be rendered normally".  (I
reject the careless remark made in the comment in the DTD, glossing
the ALT attribute as "-- short description --": this seems to me to be
very misguided, and at variance both with what it says elsewhere - i.e
this is what the TITLE attribute is for, not ALT - and with what seems
to be generally agreed best-practice in the provision of ALT texts). 

Let's please avoid the word "description" in relation to ALT texts. 
Even in contexts where it might almost seem apt, it's way too easily
misinterpreted.  And, as I say, that's what TITLE is for.  In the
present context, I'd suggest we need to say something like

  Images need ALT to serve their function.

Good work, please excuse my nit picking...

best regards
Received on Saturday, 7 November 1998 13:02:27 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Monday, 7 December 2009 10:46:58 GMT