Signifying a Complete Feed

2005-10-13 Thread Byrne Reese








Does anyone know of an extension or mechanism that can
indicate whether a given feed is a complete representation of the content it
encompasses, or is just a summary?



Here is the use case:




 Both LiveJournal and TypePad
 give users the option of not
 publishing the complete content of their posts inside their feed. In
 stead, we publish a summary of the post in the summary element.
 Thats logical enough. We see there being value however in being
 able to communicate to the reader that the contents represented by an
 entry element is not a complete representation of that entry.




Is there an existing mechanism for this anywhere? An
extension perhaps?



Byrne Reese

Manager, Platform Technology

http://www.sixapart.com/pronet/










Re: Signifying a Complete Feed

2005-10-13 Thread A. Pagaltzis

* Byrne Reese [EMAIL PROTECTED] [2005-10-13 19:40]:
 In stead, we publish a summary of the post in the summary
 element. That's logical enough. We see there being value
 however in being able to communicate to the reader that the
 contents represented by an entry element is not a complete
 representation of that entry.

Without atom:content, an atom:entry is not a complete
representation of a resource.

Is there any problem I’m missing?

Regards,
-- 
Aristotle Pagaltzis // http://plasmasturm.org/



Re: Signifying a Complete Feed

2005-10-13 Thread A. Pagaltzis

* Byrne Reese [EMAIL PROTECTED] [2005-10-13 20:55]:
 No I suppose not. But there are edge cases where people may not
 include content in a post. They may simply use the title... in
 maintaining a list of links for example. But I suppose there is
 a difference between:
 
 entry
  titlefoo/title
 /entry
 
 And 
 
 entry
  titlefoo/title
  content/content
 /entry
 
 I think that is clear enough.

That is exactly correct.

If you want to ship a complete representation, you ship an
atom:entry, and if the resource is empty, then that atom:content
is empty.

If the atom:entry has no atom:content, then that always means
that it is a partial representation only.

Regards,
-- 
Aristotle Pagaltzis // http://plasmasturm.org/



Re: Signifying a Complete Feed

2005-10-13 Thread James M Snell


Using Mark Nottingham's Feed History extension, you could do 
fh:incrementalfalse/fh:incremental


Byrne Reese wrote:

Does anyone know of an extension or mechanism that can indicate 
whether a given feed is a complete representation of the content it 
encompasses, or is just a summary?


Here is the use case:

* Both LiveJournal and TypePad give users the option of /not/
  publishing the complete content of their posts inside their
  feed. In stead, we publish a summary of the post in the
  summary element. That’s logical enough. We see there being
  value however in being able to communicate to the reader that
  the contents represented by an entry element is not a complete
  representation of that entry.

Is there an existing mechanism for this anywhere? An extension perhaps?

Byrne Reese

Manager, Platform Technology

http://www.sixapart.com/pronet/





Re: Signifying a Complete Feed

2005-10-13 Thread Graham


On 13 Oct 2005, at 8:02 pm, A. Pagaltzis wrote:



If you want to ship a complete representation, you ship an
atom:entry, and if the resource is empty, then that atom:content
is empty.

If the atom:entry has no atom:content, then that always means
that it is a partial representation only.



Point to any text in the spec that backs this up.

Graham




Re: Signifying a Complete Feed

2005-10-13 Thread Robert Sayre

On 10/13/05, Graham [EMAIL PROTECTED] wrote:
 Point to any text in the spec that backs this up.

The spec says only The 'atom:content' element either contains or
links to the content of the entry.

Any assertion that there is no other content to be had is not
testable, and therefore rightly excluded from the spec. I don't see
what harm the proposed extension could do, but it doesn't sound like
something I would implement. What's the benefit?

Robert Sayre



RE: Signifying a Complete Feed

2005-10-13 Thread Byrne Reese

If only to make the nature of a feed, or state of a feed more
deterministic and less ambiguous...?

Nottingham's Feed History achieves this objective for me. I *knew* I had
read something about this somewhere. :)

-Original Message-
From: [EMAIL PROTECTED]
[mailto:[EMAIL PROTECTED] On Behalf Of Robert Sayre
Sent: Thursday, October 13, 2005 1:51 PM
To: Graham
Cc: atom-syntax
Subject: Re: Signifying a Complete Feed


On 10/13/05, Graham [EMAIL PROTECTED] wrote:
 Point to any text in the spec that backs this up.

The spec says only The 'atom:content' element either contains or
links to the content of the entry.

Any assertion that there is no other content to be had is not
testable, and therefore rightly excluded from the spec. I don't see
what harm the proposed extension could do, but it doesn't sound like
something I would implement. What's the benefit?

Robert Sayre




Re: Signifying a Complete Feed

2005-10-13 Thread James Holderness


My understanding was that if fh:incremental was false the feed document 
should be considered a complete replacement for any previous document you 
may have received. This would be for things like top 10 lists.


I believe the question being asked here is actually about the entries 
themselves not being complete.


James M Snell wrote:
Using Mark Nottingham's Feed History extension, you could do 
fh:incrementalfalse/fh:incremental


Byrne Reese wrote:

Does anyone know of an extension or mechanism that can indicate whether a 
given feed is a complete representation of the content it encompasses, or 
is just a summary?




Re: Signifying a Complete Feed

2005-10-13 Thread James M Snell


Ah, missed that, you're right.  There is no way of indicating whether or 
not a feed is a full-content feed vs. a summary feed beyond the presence 
(or lack thereof) of the atom:content element.


James Holderness wrote:



My understanding was that if fh:incremental was false the feed 
document should be considered a complete replacement for any previous 
document you may have received. This would be for things like top 10 
lists.


I believe the question being asked here is actually about the entries 
themselves not being complete.


James M Snell wrote:

Using Mark Nottingham's Feed History extension, you could do 
fh:incrementalfalse/fh:incremental


Byrne Reese wrote:

Does anyone know of an extension or mechanism that can indicate 
whether a given feed is a complete representation of the content it 
encompasses, or is just a summary?









Re: Signifying a Complete Feed

2005-10-13 Thread James Holderness


Graham wrote:

On 13 Oct 2005, at 8:02 pm, A. Pagaltzis wrote:

If you want to ship a complete representation, you ship an
atom:entry, and if the resource is empty, then that atom:content
is empty.

If the atom:entry has no atom:content, then that always means
that it is a partial representation only.


Point to any text in the spec that backs this up.


4.2.13 The atom:summary element is a Text construct that conveys a short 
summary, abstract, or excerpt of an entry. 


To me, that implies that there exists a resource somewhere that is a more 
complete representation of what is being provided in the summary. If there 
is no atom:content element then the entry can't possibly contain that full 
representation. Surely then that would make it partial representation?


In reality though I've seen feeds with atom:content elements that wouldn't 
be considered complete by any stretch of the imagination. I've also seen 
entries with no atom:content where the full text is included in the 
atom:summary. For an Atom processor there are really no guarantees about 
anything regardless of what the spec may or may not say.


Regards
James