Le 8 déc. 2006 à 0:08, Ian Hickson a écrit :

On Fri, 8 Dec 2006, Alexey Feldgendler wrote:

Recently, "<br/>" has been brought into the common subset of HTML5 and XHTML5. That's OK because browsers currently handle "<br/>" the same in HTML and XHTML, and will continue doing so. The same for xmlns attribute
on <html>.

However, introducing <xml:base> into the common subset of HTML5 and
XHTML5 is not acceptable becasue it there woudl be markup in the common
subset that means different things for HTML5 and XHTML5 consumers:
nothing for the former, base URI specification for the latter. I don't see why would anyone want non-interoperable markup in the common subset.

I agree.

I agree too. If a something, even harmless, does not work as an author would expect it to work, it shouldn't validate. xml:lang shouldn't validate in HTML for instance because it would give authors a false sense that it correctly defines the language of the content.

So currently, the most notable things not available in the common subset are:

<base> vs. xml:base
workaround: HTTP Content-Location header
<http://www.w3.org/TR/html4/struct/links.html#h-12.4.1>

<meta http-equiv=""> vs. <?xml ?> to specify the character set
workaround: HTTP Content-Type header with charset specified.

lang vs. xml:lang
workaround: HTTP Content-Language header (although it can't switch language for different parts of a document)
<http://www.w3.org/TR/html4/struct/dirlang.html#h-8.1.2>

<noscript> and document.write()
workaround: create nodes programatically and avoid the use of <noscript>. - I'd note however that this is not a limitation specific to the common subset, but one of XHTML. It's clear that all limitations of XHTML and all limitations of HTML applies to the common subset, the interesting part is the *additional* limitations it imposes, things with are supported by both but which cannot be expressed in a cross- compatible way.

I've started a wiki page about the common subset:
<http://wiki.whatwg.org/wiki/Common_Subset>


Michel Fortin
[EMAIL PROTECTED]
http://www.michelf.com/


Reply via email to