W3C home > Mailing lists > Public > w3c-wai-ig@w3.org > January to March 2000

RE: Validation as test for basic accessibility

From: Charles F. Munat <charles@munat.com>
Date: Wed, 19 Jan 2000 13:41:07 -0800
To: "Kynn Bartlett" <kynn-hwg@idyllmtn.com>
Cc: "Web Accessibility Initiative" <w3c-wai-ig@w3.org>
Kynn Bartlett wrote:

"Don't 'we' say [make sure your code is valid] all the time?  That's where I'm a bit
confused, I think most people agree that it's a good first
step.  It simply does not meet the requirements for a priority
1 checkpoint."

No, "you" say that all the time. The amount of discussion this thread has caused shows that not everyone agrees on the importance of valid code. And when you say that it does not meet the requirements for a P1 checkpoint, what evidence to you offer to support this claim? That may be the way the WCAG currently reads, but it's not necessarily the best way. (Note that I am not suggesting that we change it, only that we should avoid making anything too sacred.)

There is a difference between the priority levels and the way we teach accessibility. Just because something is priority one doesn't mean we should necessarily teach it first (and vice versa). I don't recall anyone suggesting that valid code be moved to priority one, but that doesn't mean we can't stress it's importance (or that making the code valid makes meeting many of the other requirements easier).

"I found some of my
students were very scared... that in addition to 
learning accessible web design, they'd 'have to' learn HTML 4.0,
CSS 1, etc to have any hope of making an accessible site."

I don't see the connection. "Make sure your code is valid" and "Learn HTML 4.0" are two distinct statements. If you can get valid code from a WYSIWYG device, great! Surely you're not suggesting that it's too hard for your students to learn to use the validator.

"Most web designers, though, [do not enjoy reading technical specifications]"

I certainly don't try to teach HTML 4.0 from the specification! But I do think that we should strongly encourage people to learn HTML before attempting anything really complex. It's one thing to put up an attractive, simple page. It's another to be integrating multimedia, doing complex tables, scripting, etc.

Many web pages are effectively software. We don't blink an eye when we expect a person to learn C or C++ before writing software in it, but for some reason we think that people should be able to write HTML without learning it. I think that that's unreasonable for anything beyond a simple site, and that as the Web becomes more complex, it will become even less reasonable.

Furthermore, I didn't say (nor do I think Bruce did) that people should HAVE TO learn HTML. Only that having valid code should be stressed as a good starting point. Mostly, this is an issue that should be addressed with the makers of those WYSIWYG creators. Why do they produce invalid code? Perhaps if more people were aware of the importance of valid code, more pressure might be brought to bear on the WYSIWYG folks to fix their software so it produces valid code.

"There are many people who design web sites who have -no idea- what
HTML 4.01 is, how to make a page validate, or what the tags look
like.  However, these people need to, or want to, make accessible
web sites.  If our solution is 'you MUST learn <long list of very
technical specs>" then we have LOST all the WYSIWYG creators!'

I reiterate: no-one said anything about making anyone learn a "long list of very technical specs." On the other hand, asking people to use the validator might make them more aware of the shortcomings of their authoring software.

"I would much rather teach them sound principles of accessible
design than to require strict adherence to today's standards."

The two are not mutually exclusive. Who said anything about not teaching the "sound principles of accessible design"? You're creating an either-or fallacy here. It's not one or the other. Why not do both (substituting "encourage" for "require")?

"However, by linking [valid code and accessible design] together too
strongly we make it impossible for someone who does not speak native
HTML to create accessible web pages, and thus they won't even
bother since it can't be done."

Impossible? Strong word. Evidence to support your claim? Linking valid code and accessible design does not require one to learn code, only to be able to operate the validator. The validator could be designed to dispense advice that is less technical than it currently is. Perhaps it should suggest possibilities. For example, it could reply:

You code is not valid HTML 4.01 because:

Line 55: <center>
The <center> tag is no longer part of the HTML spec. If you are trying to center text, try replacing it with the following:


Why not a validator that also teaches?

"Validity is good, but validity is not required for accessibility,

More pronouncements from on high with nothing to back them up. Flex a little, Kynn. Is this really all that dangerous? Must we do it your way, "period"?

I suggest that we're making a mountain of a molehill here. Everyone learns in his or her own way. For some, it might be best to start with simple principles and avoid anything that even smacks of technogreek. But many people are smarter than that, and we may be missing an opportunity to interest them in delving further into the guts of the operation. Why not offer alternate approaches instead of forcing everyone into the same box?

I still think that Bruce's idea has merit. Furthermore, I think that it brings up lots of other interesting ideas, like the idea of a teaching (or at least more helpful) validator, or multiple approaches to learning accessibility.


Charles Munat,
Munat, Inc.
Seattle, Washington
Received on Wednesday, 19 January 2000 16:41:07 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 13 October 2015 16:21:07 UTC