W3C home > Mailing lists > Public > www-validator@w3.org > March 2010

Re: Validation bug - XHTML 1.1 and XHTML+RDFa

From: Jukka K. Korpela <jkorpela@cs.tut.fi>
Date: Mon, 15 Mar 2010 20:11:32 +0200
Message-ID: <C499160443C84A1C86FED04116D7C184@JukanPC>
To: <www-validator@w3.org>
Leif Halvard Silli wrote:

> XHTML 1.1 doesn't cover the lang attribute - it only supports
> xml:lang.

Indeed. That's one of the many reasons for not using XHTML 1.1 (for WWW 
documents at least).

> But when I actually tried to validate a page with @lang,
> then it validated. How come?

It's of course a bug, but I'm afraid things are worse:

> I have heard that XHTML 1.1. will soon be redefined to support @lang
> (which is great!).

I have heard that XHTML 1.1 is dead and really wasn't much alive ever (which 
is great). I don't know about any activities around it.

> Is Validator.w3.org anticipating this?

Who knows? It is still a bug. It is a serious bug in a program called markup 
validator that it claims a document to be valid when it is not and, worse 
still, if this is intentional and based on some guess of what users "really 
want".

Maybe it's not an intentional bug as such but as a byproduct of the 
guesswork based on the public identifier. The bug disappears if you use

<!DOCTYPE html SYSTEM
    "http://www.w3.org/TR/xhtml11/DTD/xhtml11.dtd">

so clearly there is some public identifier magic in the air.

The validator at http://www.htmlhelp.com/validator hasn't got this bug. 
There has probably been very little if any development around that 
validator, and this seems to be a good thing. There isn't really much you 
can do better in SGML or XML validation, and the history of thje W3C 
Validator seems to prove this. If you wish to make your software do 
something completely (or at least essentially) different from such 
validation, that might be a very good idea, and there are many checks that 
could be much more useful than markup validation, but that's a different 
issue.

> If there is some kind of anticipation going on, then that's OK for me

It would be rather absurd to modify XHTML 1.1 in one of the few features 
where it considerably differs from XHTML 1.0, taking it back to 1.0 
position, and still call the result XHTML 1.0. Public identifiers of 
document types should be immutable: when you change a document type, you 
define a new public identifier (if you wish to use one).

-- 
Yucca, http://www.cs.tut.fi/~jkorpela/ 
Received on Monday, 15 March 2010 18:12:57 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Wednesday, 25 April 2012 12:14:40 GMT