Graham wrote:
1. Why are these two statements in the same paragraph?
The content of an atom:id element MUST be created in a way that assures uniqueness; it is suggested that the atom:id element be stored along with the associated resource.
Storage assures persistence but does nothing to improve uniqueness. I'd recommend that the second statement be moved to the end of the paragraph before, and the remaining statement be found a home elsewhere.
That sentence is an oblique way of saying "don't use your auto-incrementing primary key or your title-based permalink to identify entries". Suggestions welcome.
2. This would leave room to discuss uniqueness in more depth, which currently isn't discussed. Is it acceptable to have a feed and entry use the same identifier? Does it imply some relationship? Is it acceptable to use a domain name I don't control (this would obj=viously threaten uniqueness), etc etc. Probably too late to get consensus on changes here, though, so I won't suggest any.
I agree that it's too late. Those are hard questions.
3. Regarding dereferencing:
Though the IRI might use a dereferencable scheme, Atom Processors MUST NOT assume it
can be dereferenced.
This is nice, but I'd prefer it to also say something like "Atom Processors MUST NOT make any assumptions about the relationship between the feed or entry and the dereferenced object". This is currently implied by the absence of any discussion, so I don't think it's a functional change.
No matter how much we might wish it to be so, we can't outlaw that sort of thing.
Robert Sayre
