Re: [apps-discuss] [rest-discuss] Re: Feedback on draft-wilde-profile-link-00

[Julian Reschke <[e-mail]>]

On 2012-04-12 19:26, Erik Wilde wrote:
> hello julian.
>
> On 2012-04-12 09:52 , Julian Reschke wrote:
>> Maybe we need to switch to examples to get a clearer picture.
>> With "profile" I understand that the link *target* is a URI describing
>> the profile, such as <http://www.w3.org/2006/03/hcard> which describes
>> that hcard profile.
>
> no, the 'profile' link target URI *identifies* the profile in use, and
> it MAY point to a description (and hopefully will), but it does not have
> to do that. at that level it works exactly the same as XML namespace

Yes, sorry for being not precise.

> URIs: its definition is to be an identifier, but if you want to be nice,
> then you actually link to some description, but that's entirely up to
> you. here's what the draft currently says:
>
> "Profiles are identified by URI, but as with for example XML namespace
> URIs, the URI in this case first and foremost serves as an identifier,
> meaning that the presence of a specific URI has to be sufficient for a
> client to assert that a resource representation conforms to a profile."
>
>> If you want to use the profile link relation to declare that something
>> is a an AtomPub server, on which resource would you set the link
>> relation?
>
> the feed served by the atompub-enabled server would contain such a
> 'profile' link on the feed level, indicating to a client that it can
> POST to that feed's URI to create a new collection member.

OK. I don't have a problem with that. The one thing to consider is, as 
others have pointed out, whether it would make sense to use "service" 
instead and let that point back to the same URI (assuming we have a 
proper definition of "service" :-)

Best regards, Julian