Re: rel values

[Danny Ayers <danny.ayers@xxxxxxxxx>]

On Sat, 9 Oct 2004 08:38:58 -0700 (PDT), Dare Obasanjo <kpako@xxxxxxxxx> wrote:
> 
> I don't think anyone has ever complained about the
> syntax 

I got the impression the first post in this thread was doing just that.

so your proposed solution does nothing to
> change the problem. The problem being, exactly what is
> the semantics of a link so consumers of Atom feeds can
> know how to process them in a uniform manner.

That's not difficult -

"The link construct defines a binary relationship between the parent
of the element in which it appears, and the resource identified in the
atom:href attribute. The nature of that relationship is detemined by
the name of the element containing the atom:href attribute. If that
name is in the Atom namespace, then a consumer MUST interpret it and
behave according to the definition of that term in this specification.
Otherwise the consumer MAY follow whatever definition is provided for
the the term in the namespace in question. If the consumer has no
knowledge of the term then it MAY ignore it or present the data to the
end user in any manner deemed appropriate".

Examples (something like):

<atom:entry>
   <atom:id>http://example.org/blah</atom:id>
   <atom:service atom:method="edit" atom:href="http://example.org/editme"; />
...
</atom:entry>


This says: the edit endpoint using the Atom Protocol for the resource
http://example.org/blah is http://example.org/editme.

<atom:entry>
   <atom:id>http://example.org/blah</atom:id>
   <foaf:depiction atom:href="http://example.org/veg.jpg"; />
...
</atom:entry>

This says: http://example.org/veg.jgp is a foaf:Image which depicts
the resource http://example.org/blah

Cheers,
Danny.

-- 

http://dannyayers.com