Posted by Andy Dingley on 02/04/06 16:02

On Fri, 03 Feb 2006 21:31:15 -0500, John Salerno
<johnjsal@NOSPAMgmail.com> wrote:

>Anyhow, after finding several sites very much against XHTML (but some in
>favor), and especially after reading: http://hixie.ch/advocacy/xhtml it
>makes a lot more sense to me now why XHTML is considered more or less a
>useless language.

XHTML is demonstrably very far from useless.

I work mainly on content-management systems, not hand-authoring directly
to the web. My page content might move through half-a-dozen separate
processes before finally reaching the web. Through all of these steps,
using XHTML is a _massive_ benefit to me, rather than HTML. There is
just no question of this - XML processing tools are cheap, easy and
powerful, everything that SGML failed with completely.

For the final output, I can transform to HTML output and in some ways
this is even easier than XHTML (it's hard to generate good Appendix C
XHTML from most XSLT tools)

Appendix C XHTML is a kludge, but it's a kludge that works on the web
and allows XHTML to be served in a way that's at no disadvantage
compared to HTML. Now if I'm already going to be using XHTML
internally, then who benefits from pushing it into another yet format
just to serve out ? I need a good argument for going back to HTML over
Appendix C.


Hixie's position is pure sophistry. He constructs a valid argument
against a proposition no-one is really advocating (Authoring XHTML so it
can be treated as either XML or HTML simultaneously) then he attacks the
real situation (serving XHTML to existing HTML browsers) on the basis of
spec subtleties that no credible browser ever supported. I'm still
waiting for a screenshot of a browser demonstrating the "infamous
SHORTTAG bug"

I particular he creates two notable straw men
* The "xmlns" attribute is invalid HTML4.
* The XHTML DOCTYPEs are not valid HTML4 DOCTYPEs.

Why should we care that the xmlns attribute is invalid ? Why should we
care about validation at all ? Validation is useful because it's
objective and simple (you're either valid or not), but equally there is
no benefit accruing from validation _itself_, only from achieving a
standard of "objective compatibiity" that is most easily achieved by
demonstrating validity. In the case of xmlns though, the
well-established rules on ignoring unknown attributes can handle things
perfectly adequately.


As th the doctype validity issue, then this is bogus. The XHTML doctype
is perfectly valid, it's merely different. It's also well-established
for some years now and widely understood by browsers (to the extent that
browsers do anything useful with a doctype anyway).

If we take Hixie's own position of "Ivory tower SGML purist who hasn't
even noticed the M$oft barbarians at the gate", then doctypes have
always been flexible and extensible by SGML's rules. Now in web terms
this was a bad approach and never really worked (despite the number of
late-'90s HTML editors that tried). A custom doctype gains you nothing
on the web, it's not used by an SGML parser to extend HTML, and the very
best you can hope for is that a doctype is seen as one of a handful of
magic identifier strings that the browser recognises. Perhaps it's a
sadness that the web chose to never use this feature and go down this
route (personally I don't think so, but that's for another thread).
However I do not have to listen to arguments against varying a doctype
in a respectable from someone who is simultaneously castigating me for
SHORTTAG ! Take your pick - the ivory standards or the real web
practices - you can't pick and choose whichever happen conveniently to
support your own position.


Hixie is asking me to throw away the processing capabilities of XML in
favour of pleasing the tiny handful of SGML-anoraks who even understand
what the problem is. This is no bargain.

Meanwhile the rest of the world sees MS Office and Dreamweaver as
appropriate HTML authoring tools, despite their absolutely glaring
holes. The enemy here is bad and bogus markup with no structure
whatsoever, not XHTML.

[Back to original message]