... taking XHTML seriously ...

**esteban** · 05-03-2006, 08:08 AM

Hi folks. I was skimming through some recent threads and I saw this:

Originally Posted by mwinter

... There is absolutely no point writing malformed pseudo-XHTML tag-soup. Indeed, doing so will lead to a nasty shock if such a document were to be served to a conforming XML processor. It should be valid, XHTML Strict or nothing (and only XHTML 1.0 if served as HTML).

Of course, as I've written several times in the past, serving XHTML as HTML is pointless at the best of times. However, one cannot claim to be 'moving forward' by using XHTML (not that such an argument holds any water) if it's exactly the same kind of rubbish found commonly with HTML.

I would appreciate it if folks could elaborate on this.

As I read it, the "transitional" doctypes are less-than-desirable?

As always, thanks in advance

.

**djr33** · 05-03-2006, 08:21 AM

I'm also interested in learning more about this

**Twey** · 05-03-2006, 08:30 AM

The point of the standards and DOCTYPEs is to encourage well-formed code. Transitional DOCTYPEs go against this principle as a "stepping stone" from an old DOCTYPE (or possibly no DOCTYPE) to a new one, and are not intended as a permanent measure. There should now be no need to use Transitional DOCTYPEs since sites should be using up-to-date code already. One should certainly never intentionally design a new site based on a Transitional DOCTYPE.

What Mike was arguing against there was XHTML, however, not loose DOCTYPEs in general. XHTML is currently not supported by IE (it will attempt to parse it as malformed HTML if sent as text/html [which is only allowed with 1.0, not 1.1], or offer the user the choice of what to do with it if correctly sent as text/xhtml) and is rendered more slowly and less efficiently than HTML in other browsers. Therefore, XHTML is not currently viable for general-purpose use on the Web, as Mike says.

**esteban** · 05-04-2006, 07:05 AM

Originally Posted by Twey

The point of the standards and DOCTYPEs is to encourage well-formed code...

What Mike was arguing against there was XHTML, however, not loose DOCTYPEs in general. XHTML is currently not supported by IE (it will attempt to parse it as malformed HTML if sent as text/html [which is only allowed with 1.0, not 1.1], or offer the user the choice of what to do with it if correctly sent as text/xhtml) and is rendered more slowly and less efficiently than HTML in other browsers. Therefore, XHTML is not currently viable for general-purpose use on the Web, as Mike says.

Wow. I am ashamed to admit that I thought IE supported XHTML.

Thanks for the info

Given the current state of things, do you think HTML 4.01 Strict is an ideal doctype for general-purpose use on the web?

If the answer is too involved, throw some relevant links / articles my way

. I don't mind reading.

EDIT: I found this Sending XHTML as text/html Considered Harmful, Author: Ian Hickson.

UPDATE: OK, after some preliminary research, it seems like HTML 4.01 Strict is the way to go...

And here I was, thinking that if I wasn't using XHTML then I was behind the times .... I'd still love any advice you folks have, of course...

**mwinter** · 05-04-2006, 12:22 PM

What Twey wrote is the condensed version.

Originally Posted by esteban

Wow. I am ashamed to admit that I thought IE supported XHTML.

You aren't alone. MSIE can only process HTML. The upcoming version (7) will be the same.

Given the current state of things, do you think HTML 4.01 Strict is an ideal doctype for general-purpose use on the web?

Yes, and, if well written, can easily be converted to XHTML (should it become viable). However, someone predicted in Usenet recently that, if anything, XHTML 2.0 may become the future markup language. It is a completely different animal compared to XHTML 1.x (it's not backwards-compatible, for a start). Only time will tell.

If the answer is too involved, throw some relevant links / articles my way. I don't mind reading.

The debate has been done to death on Usenet. The Google archives contain several discussions of the subject.

I'd still love any advice you folks have, of course...

As you'll probably read elsewhere, there's no problem using XHTML and XML for data storage. Indeed, a CMS designed for XML (don't use one that was retrofitted, though) can offer a lot of flexibility. The argument against X(HT)ML is its use for serving to clients: except in rare circumstances (like using MathML), there's zero benefit.

Mike