[Irl-dean] Opinions: HTML Transitional and WCAG Double-A

[Barry McMullin]

On Wed, 10 Jan 2007, Joshue O Connor wrote:

> So have we reached a con census? Can we say it is better to recommend
> use of an HTML strict doc type for most web development tasks?

I would say "yes"...

> What about strict doc types for XHTML?

I think XHTML vs. HTML should be treated as a completely separate
question from strict versus transitional.  (That said: I see a
lot of value in using XHTML on the server side - particularly to
integrate with any more general XML-based tools and systems; but
I have seen virtually no real-world applications yet where there
was *any* advantage, on accessibility or any other grounds, to
shifting to XHTML on the client side. So my advice is to use
XHTML internally on the server side, by all means - along with
XML, XSLT, PHP, perl, tcl, SQL and anything else that tickles
your fancy; but to serve *only* HTML to the client side... XHTML
on the client side is, as yet, nothing more than a fashion
statement, which delivers no extra value to users and which, in
at least some scenarios, actually causes definite problems, both
accessibility and otherwise. Roll on HTML 5!)

> Will using a strict doc type for either HTML and XHTML have a
> negative impact on legacy user agents ...

Apparently only in the relatively specialised cases which we have
managed to identify already (e.g., ordered lists with fine
tuned/out-of-sequence numbering).  The main issue here is not so
much HTML strict per se, but support for CSS (which is what
removes the need to use the transitional features of HTML).  We
all recognise that good, solid, CSS support has been very slow to
arrive (particularly in IE).  But the "To Hell with Bad Browsers"
manifesto dates from February 2001!  I'm generally a fairly
conservative kind of guy, but, in the long term interests of
accessibility, I for one, am ready to move on and say designers
should now work on the assumption of CSS 2.1 (which essentially
allows HTML transitional to be abandoned); and only back off from
that, if at all, on foot of a clear and demonstrated problem
among their real user base...

> or if we recommend to other developers to use strict DTD's for
> either markup languages will it in fact be the push they need to
> _really_ separate content/presentation/behaviour?

Well, I can't vouch for that; but it might at least shift the
discussion from the absolutely trivial, "me too", "XHTML (um
... transitional)" bandwagon, which, to my thinking, is
delivering essentially nothing for accessibility...

Best - Barry.