> On 10/31/05, Luke Arno <[EMAIL PROTECTED]> wrote:
> 
>>I think that Atom syntax provides us with what we need.
>>
>>Not only do I think that we can deal in valid entries, 

More detail on your thinking please.

1. IDs. The protocol operations can in principle change, depending on
whether we allow/deny clients to generate ids. Check the archives, I
brought this up already plus ascii art (thread "Ways of uploading
content").

2. Atom Elements. At the moment the interesting elements are:

 atom:id
 atom:updated
 atom:link

the first two are specced as "server controlled" in 06. In "8.1 Creating
resources with POST" the XML example was changed to be the minimal
(invalid) atom that could be sent.

3. Invalid Atom. Please note this text in 8.1, para4:

"Clients MAY POST invalid Atom for initial resource creation -
specifically the id and link elements MAY be omitted."

That's new text since 05 and everyone here is invited to focus on it for
discussion. It's been added because it is the case, and has been for
some time - 06 makes it explicit that we are very likely underspecified
and almost certainly in the business of sending invalid atom around.
It's unclear whether this invalidity is a format or a protocol
consequence, or both.

cheers
Bill

Reply via email to