Re: New article: Indicating the language of a link destination

On 06/09/2013 15:46, Anne van Kesteren wrote:
> On Fri, Sep 6, 2013 at 3:27 PM, Richard Ishida <ishida@w3.org> wrote:
>> The article discusses some of the pros and cons for signalling the language
>> of a page which a link points to, if that page is not in the same language
>> as the current content. It also looks at how people have done this in the
>> past using the hreflang attribute.
>
> I wonder why we still have this attribute. It seems if the only
> practical benefit is as a CSS styling hook it might be better to place
> it in content as that would be more accessible too. And by having it
> in the language it will waste people's time wondering what to do with
> it. Any reason hreflang="" is still in HTML, Ian?

I wonder the same thing, though i have seen people using it in the way 
described in the article.  It's a bit hamstrung for that usage, though, 
given that it only supports a single language value.

After discussions, I believe iirc that the developers of XHTML2 were 
planning to redefine the hreflang attribute so that it actually would 
send information with the Accept-Language header requesting a particular 
language version of the resource to be fetched. I'm not sure how often 
that would get used, but it did at least seem to be slightly more useful 
than the current defininition of the attribute.

I agree with the HTML5 spec, btw, that hreflang should not be used to 
provide remote declarations for the language of the targeted resource - 
people should do that in the resource itself (for reasons very similar 
to the cons in the article about what happens if circumstances change).

There may be other uses of hreflang that i'm missing though.

my 2p,
RI


-- 
Richard Ishida

Received on Friday, 6 September 2013 15:06:12 UTC