|
|
Posted by Toby A Inkster on 05/21/07 11:36
Andy Dingley wrote:
> On 20 May, 08:55, Toby A Inkster <usenet200...@tobyinkster.co.uk>
> wrote:
>
>> Actually, by virtue of their support for XML+CSS, most browsers can render
>> the majority of XHTML2 already if you provide them with an appropriate
>> stylesheet for the new elements.
>
> That's more of a bogosity than Appendix C!
What you quoted is bogus, yes, but that's not what I said. I added that
client-side scripting is needed to emulate some features.
> * It requires a non-standard CSS extension to generate links, as
> standard CSS can't create a href attribute on an <a>
You don't *need* to use CSS to create an href attribute for <a>. XHTML2
for a link is:
<a href="...">...</a>
which is already supported in browsers, provided you use an XHTML
namespace. XHTML2 makes the href attribute "special" though, instead of
the <a> element. So the following is also a valid link:
<cite href="...">...</cite>
which is why I specified that client-side scripting is needed -- to read
the href attribute of non-<a> elements, and either dynamically create <a>
elements to provide link functionality, or set onclick attributes to
emulate it.
> * It abandons the little semantics that HTML (any sort) already has in
> favour of a semantics-free XML that the browser no longer recognise as
> being HTML-like. This is worse than the tag soup position.
XHTML2 is not semantics-free XML. It is capable of expressing semantics far
better than existing HTML can. Currently though, far fewer tools are
available to deal with it.
I didn't claim that using XHTML2 today is a good idea -- just that
existing user agents (Opera and Mozilla I've tried, but I'd guess that
KHTML/WebCore-based agents too) are able to deal with it, providing the
author includes some CSS and Javascript to smooth over the differences.
I do know what I'm talking about -- my CMS <http://demiblog.org/> has
experimental support for current drafts of XHTML2 *and* HTML5. I wouldn't
advise using them on production sites, but if someone's committed to using
them, then demiblog won't get in their way. (For what it's worth, it also
outputs standards-compliant HTML 4, HTML 4.01, XHTML 1.0, HTML 1.0 Basic,
HTML 1.0 Print and XHTML 1.1. Experimental support for ISO HTML is also
available and works pretty well. All the markup produced by the CMS
validates, but plugins and user-generated content may break validation.
There's not really any way to stop that.) I have produced content in
XHTML2, and it *can* be made to work, today, if you're determined enough.
--
Toby A Inkster BSc (Hons) ARCS
http://tobyinkster.co.uk/
Geek of ~ HTML/SQL/Perl/PHP/Python/Apache/Linux
Navigation:
[Reply to this message]
|