Re: LDP Agenda for December 17, 2012, with a list of issues to be closed

On 15/12/12 23:52, Steve Battle wrote:
>
> On 15 Dec 2012, at 21:31, Ashok Malhotra <ashok.malhotra@oracle.com
> <mailto:ashok.malhotra@oracle.com>> wrote:
>
>>
>> On 12/15/2012 10:41 AM, Arnaud Le Hors wrote:
>>>
>>> So, to clarify, the model applications can handle on their own is a
>>> basic aggregation model where the server isn't expected to do
>>> anything special - like deleting member resources when the container
>>> is deleted. As of Lyon, we agreed to define the composition model in
>>> LDP and left the aggregation model to be proposed by those who feel
>>> the need for it. We have yet to see a concrete proposal.
>>
>> PROPOSAL:
>> Add an attribute to Container called DeleteMembers.  Boolean. Default

Ashok:

By "Attribute" you mean a property of the container?
Or is this per container entry?
Or a platform managed aspect not in the RDF?

>> = YES
>
> -0.5
>
> I really don't like this. If this were false then why use a container in
> the first place. Just use aggregation and describe the collection in
> RDF. The LDP spec doesn't need to be concerned with what we can say in RDF.
>
> Steve.

While I like this as a principle, there are some details ...

* LDP-C have paging. RDF collections and/or containers in a LDP-R do not.

* Different member entry creation mechanisms (adding to a RDF list or 
RDF Seq is tricky at the best of times).

A flag on a LDP-C to say whether it is an aggregation or a management 
containment relationship is, in effect, having two types - an LDP-C(agg) 
and LDP-C(mgt) that share paging, adding new members, etc

	Andy

Received on Sunday, 16 December 2012 16:30:16 UTC