RE: Additional thoughts on language specification

> Roberto Scano:
> Here there are some tests:
> [fix] http://www.w3.org/International/tests/results/lang-decl

Thanks Roberto, this is extremely helpful.

>From the discussion so far, I think we can go ahead and and have two
sufficient techniques: the lang-attribute of the HTML element (if HTML 4 is
in the baseline) 
and the xml:lang attribute of the HTML element (if XHTML 1+ is in the
baseline). Within the last technique we could recommend using the
lang-attribute as well for backward compatibility, although that would not
be required for conformance.

In addition to specifying the language using lang or xml:lang, we can have
two further optional techniques to indicate the primary language in the HTTP
headers or in the META tag but these techniques are not sufficient to
conform. Do we have a place for HTTP header techniques? We might need an
orphaned techniques document to put techniques that don't have a home yet. 

We still have no sufficient techniques to identify the primary language for
HTML 3.2 or lower. I think that highlights a fundamental problem with the
current WCAG spec: it does not allow the use of technologies that have no
features to indicate the primary language. Do we really want to claim a site
can't be even minimally accessible unless you use a technology that allows
you to indicate the primary language? 

I'm not sure yet if I'll have the time to write/spit polish these techniques
today or tomorrow (before my holiday) but I'll do my best. 

Yvette Hoitink
Heritas, Enschede, the Netherlands
E-mail: y.p.hoitink@heritas.nl
WWW: http://www.heritas.nl 

Received on Tuesday, 13 September 2005 14:50:54 UTC