Re: Last Call: draft-nottingham-http-link-header (Web Linking) to Proposed Standard

Noah Slater wrote:
> Ah, okay.
> 
> I got very excited for a second here because I thought you were saying that
> Firefox understands the Link header with rel="stylesheet", but I just tested it
> on one of my pages and it doesn't seem to work:
> 
>   http://tumbolia.org/quote/zen
> 
> I am sending this:
> 
>   Link: </style>;rel=stylsheet
> 
> Should this work? Am I doing something wrong?

It works in some cases; I imagine it wants a mime type as well. See 
<http://www.hixie.ch/tests/adhoc/http/link/>.

>>> The hierarchical position of each link can be found by parsing the URI.
>> Well, that won't always be the case. In the case where the URL hierarchy
>> happens to be the actual hierarchy, you really don't need the link
>> relation (as demonstrated by the Firefox addon).
> 
> It is not as simple as that:
> 
>   <link rel="up alternative" type="application/atom+xml" href="/index.atom">

"up alternative"? I hope these are two different link relations?

>   <link rel="up" hreflang="en" href="/dir-a/index.en">
> 
>   <link rel="up" hreflang="de" href="/dir-a/index.de">
> 
> In the above cases:
> 
>   * You could not infer the "up" relationship from the URI because they have
>     character data after the final "/" character, making it otherwise ambiguous
>     whether they can be considered direct "parents" of the current resource.
> 
>   * You can, however, infer the hierarchical position from the URI after being
>     told that they are parents in a document hierarchy.
> 
> The only possible time I can think of needing "up up" would be when each parent
> link was outside of the URI hierarchy of all the other links. This seems like it
> would be such an uncommon thing. Is it worth specifying for?

No, I don't think it's worth it.

BR, Julian

Received on Sunday, 30 August 2009 17:53:08 UTC