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

Fwd: Re: Check existence of class names

From: Siegfried Gipp <siegfried@rorkvell.de>
Date: Wed, 18 Aug 2010 10:55:32 +0200
Message-ID: <4C6BA004.5070607@rorkvell.de>
To: www-validator@w3.org


-------- Original-Nachricht --------
Betreff: Re: Check existence of class names
Datum: Wed, 18 Aug 2010 10:51:40 +0200
Von: Siegfried Gipp <siegfried@rorkvell.de>
An: chukharev@mail.ru

Hi,

just to clarify this somewhat further, here another example.

There existe (among others) the Dublin Core initiative
http://dublincore.org/

That is a consortium defining some class names for some special
purposes. In the old days these where just and only values to be set to
the <meta> element. But now in xml and xhtml there are more options.

So first, if you want to use these class names, you add a namespace
declaration:

xmlns:dc="http://purl.org/dc/elements/1.1/"

(see http://web.resource.org/rss/1.0/modules/dc/)

Now you can use class names like

<span class="<dc:subject>Discussion about non-existent classes</span>

Now this class name may be verified indeed. How? Well, at the url
defined in the namespace declaration you find an rdf file not only
listing all valid values, but also their proper usage and meaning. So
using these class names they could indeed be verified. They could be
verified against typos, too.

But a) that has no impact on class names without such a namespace
prefix, and b) in plain html it is not possible to use namespaces. So in
plain html there is indeed no possibility to reliably validate any value
of any class attribute. There simply does not exist any formal
definition of right or wrong. For DC there is.


Received on Wednesday, 18 August 2010 08:53:14 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Wednesday, 25 April 2012 12:14:43 GMT