Yes, the issue did exist before given that the linked atom feed could easily 
have multiple atom:entry elements with the same atom:id value as allowed by 
rfc4287. It is a simple matter to answer whether there can be multiple distict 
child sets or a single logical child set and to answer whether membership in 
that set is based on instance or identity. 

- James  
Sent from my Verizon Wireless BlackBerry

-----Original Message-----
From: "Nikunj R. Mehta" <[email protected]>

Date: Wed, 10 Jun 2009 11:15:04 
To: James M Snell<[email protected]>
Cc: atom-syntax Syntax<[email protected]>
Subject: Re: New Version Notification for draft-divilly-atom-hierarchy-02


On Jun 10, 2009, at 10:55 AM, James M Snell wrote:

> Thank you for the changes. Another issue that needs to be looked at  
> (which existing in the previous version of the draft as well)

This issue did not exist in the previous draft since only one link  
with @rel value down and @type application/atom+xml was possible.

> is how a processor can reconstruct the logical set of children for  
> an entry... that is, for instance, if the set of child entries  
> contains multiple atom:entry elements with the same atom:id value,  
> are those viewed as separate children or is only the entry with the  
> most recent atom:updated value included in the set of children?

We have two choices - either constrain cardinality and obtain an  
unambiguous logical set of children or remove the cardinality  
constraint and leave it to applications to interpret multiple sets or  
subsets. I don;t see how we can get a third option. BTW, the threading  
extension uses the second approach and §2.2 as well as §2.3 of  
hierarchy I-D explain what an Atom processor does with multiple  
occurrences of the down and up link, respectively.

> Further, if multiple "down" links are specified, each pointing to  
> separate Atom feeds with separate entries with distinct atom:id  
> values, is that to be viewed as two separate child collections or a  
> single logical combined collection? I would recommend the latter.

As I explain above, this choice is infeasible. How is one to reconcile  
an HTML linked resource with an entry with a feed and an image all  
through @rel=down? I would be interested in learning of a proposal to  
handle this kind of a situation.

>
> Nikunj R. Mehta wrote:
>> Based on feedback, I have simplified the I-D to:
>> In-line extensions moved to draft-mehta-atom-inline
>>      Removed down-tree and up-tree relations
>>      Removed cardinality restrictions on up and down links
>>
>> HTML: http://tools.ietf.org/html/draft-divilly-atom-hierarchy-02
>> Text: http://tools.ietf.org/id/draft-divilly-atom-hierarchy-02.txt 
>> <http://tools.ietf.org/id/draft-mehta-atom-inline-00.txt 
>> >
>> Diff: http://tools.ietf.org/rfcdiff?url2=draft-divilly-atom-hierarchy-02.txt
>>
>> I am also tracking open issues about this I-D publicly at 
>> http://code.google.com/p/atom-ext/issues/list 
>> . The source for this I-D is also available, if you are interested.
>>
>> Looking forward to comments on the I-D.
>>
>> Nikunj
>> http://o-micron.blogspot.com <http://o-micron.blogspot.com/>
>>
>>
>> Begin forwarded message:
>>
>>> *From: *IETF I-D Submission Tool <[email protected] 
>>> <mailto:[email protected] 
>>> >>
>>> *Date: *June 9, 2009 5:33:57 PM PDT
>>> *To: *[email protected] <mailto:[email protected]>
>>> *Cc: *[email protected] <mailto:[email protected]>
>>> *Subject: **New Version Notification for draft-divilly-atom- 
>>> hierarchy-02 *
>>>
>>>
>>> A new version of I-D, draft-divilly-atom-hierarchy-02.txt has been  
>>> successfuly submitted by Nikunj Mehta and posted to the IETF  
>>> repository.
>>>
>>> Filename: draft-divilly-atom-hierarchy
>>> Revision: 02
>>> Title: Hierarchy Relations for Atom
>>> Creation_date: 2009-06-09
>>> WG ID: Independent Submission
>>> Number_of_pages: 7
>>>
>>> Abstract:
>>> This specification defines link relations for hierarchical  
>>> navigation
>>> among Atom feeds and entries.Editorial Note
>>>
>>> To provide feedback on this Internet-Draft, join the atom-syntax
>>> mailing list (http://www.imc.org/atom-syntax/) [1].
>>>
>>>
>>>
>>> The IETF Secretariat.
>>>
>>>
>>

Reply via email to