Re: MUST a collection be returned as an Atom feed?

[Joe Gregorio <joe@xxxxxxxxxxxxxx>]

On Mon, Dec 28, 2009 at 12:08 PM, Jan Algermissen
<algermissen1971@xxxxxxx> wrote:
>
> On Dec 28, 2009, at 5:49 PM, Joe Gregorio wrote:
>
>>
>> On Mon, Dec 28, 2009 at 11:32 AM, Jan Algermissen
>> <algermissen1971@xxxxxxx> wrote:
>>>>
>>>>
>
> [...]
>
>>>
>>> Rephrasing the question:
>>>
>>> Whay is the sentence "Collections are represented as Atom Feeds" in RFC
>>> 5023?
>>>
>>> Is it a MUST requirement for AtomPub servers?
>>
>> Yes.
>
> That is what I understood from the spec. Thanks.
>
> What is your reaction to the following?
>
> Given it is a MUST requirement, it can be said that clients can make the
> assumption to receive an Atom feed document as the response to a GET request
> to an AtomPub collection resource. Yes?

No. An Atom representation MUST be present, but the client should still send
an appropriate Accept header since there might be multiple representations
available. See http://tools.ietf.org/html/rfc5023#section-4.4

> If so, that means the client can make assumptions about the next available
> links (state transitions) after receiving the response to the GET request.
> E.g. it knows there will be links to member resources in the feed. Yes?

Just like when I retrieve an HTML page I know that it might contain links
to other resources, but that's as far as my 'knowing' goes. I don't 'know'
what the actual URIs are, or even if they are going to be there, i.e. it
could be a feed with 0 entries.

   Thanks,
   -joe

> If so, doesn't that violate RESTs hypermedia constraint because RFC 5023
> enables the client to make design time assumptions about the state machine
> provided by the service? Shouldn't the client allways just discover the
> available transitions instead of assuming them?
>
> Jan
>
> (Please excuse me if you think this is nit picking. I am not just trying to
> be difficult)
>
>
>
>
>
>
>
>>
>>>
>>> (Or a SHOULD or a hint to client developers?)
>>>
>>>
>>> Or, rephrasing again:
>>>
>>> Could the sentence "Collections are represented as Atom Feeds" be removed
>>> from RFC 5023 without consequences?
>>
>> No it could not be removed, RFC 5023 is the Atom Publishing Protocol, it's
>> about
>> publishing using Atom Feed and Entry documents.
>>
>> You could remove it and let the client try to guess the media-type
>> of the resources, or work with a range of media-types, such as JSON, HTML
>> and Atom, which might be an interesting experiment, but it wouldn't be
>> AtomPub.
>>
>>  Thanks,
>>  -joe
>>
>>>
>>>
>>> Why is it in there at all?
>>>
>>>
>>>
>>> Jan
>>>
>>> P.S. I am using that sentence as a stripped down example of a point I am
>>> trying to make. No intention to discuss/question RFC 5023 in particular.
>>>
>>>
>>>
>>>
>>>
>>>> Nikunj Mehta
>>>> http://blog.o-micron.com
>>>>
>>>>
>>>>
>>>>
>>>>
>>>
>>> --------------------------------------
>>> Jan Algermissen
>>>
>>> Mail: algermissen@xxxxxxx
>>> Blog: http://algermissen.blogspot.com/
>>> Home: http://www.jalgermissen.com
>>> --------------------------------------
>>>
>>>
>>>
>>>
>>
>
> --------------------------------------
> Jan Algermissen
>
> Mail: algermissen@xxxxxxx
> Blog: http://algermissen.blogspot.com/
> Home: http://www.jalgermissen.com
> --------------------------------------
>
>
>
>