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

[Julian Reschke <[e-mail]>]

On 2012-04-12 18:06, Erik Wilde wrote:
> hello.
>
> On 2012-04-12 08:49 , Julian Reschke wrote:
>> On 2012-04-12 09:12, Jan Algermissen wrote:
>>> (BTW, seems the registry is broken there, because service cites
>>> atompub as its spec - which it is not).
>> Indeed, let's blame Mark for this :-)
>
> yup, that occurred to me a while ago when i was looking for a spec for
> the "registered" 'service' relation and there was none, but back then,
> nobody seemed to think that was an issue that needed fixing. i think it
> would be good to have an actual spec (just out of curiosity it would be
> interesting to know how this value was able to creep in there without
> being spec'ed or reviewed a la rfc5988), and since i am in spec writing

Well, because it was in the Atom link relation registry before RFC 5988 
was approved.

> mode these days, i could easily get a 'service' one started. however,
> since this one would likely update rfc5023, i'd suggest to maybe also
> (if we can get agreement around this) include a 'profile' URI, assuming
> that there is some agreement on the 'profile' concept and that it would
> make sense for signaling that "an atom feed is following the atompub
> profile".
> ...

Not sure about that.

It seems you are using "profile" in a very different way then it was 
used in HTML4; is this really a good idea?

Best regards, Julian