Bill de hOra wrote:
But Bill, it's not two different feeds, it's two different views of
the same feed -- an archive view, and a "latest entries" (a.k.a.
"sliding window") view. Thus the shared atom:id IS "globally unique
on a per my feed basis", just not on a per feed DOCUMENT basis.
What breaks if they have two different ids?
Perhaps nothing "breaks", but it's just a little odd to have the same
entries originating from two different feeds. I presume neither
document is going to use an atom:source element to show that they're not
the originator of the entries, so it looks like you have two feeds and
are erroneously using duplicate IDs for their entries.
> Put another way, why are two URLs required, but only one id? Are views
> (for any definition of "view") really a special case?
Two URLs: so that people can choose to get either the archive of or the
live end of the feed.
One ID: because they're the same feed.
And by the way, is the archive document one monolithic document
containing everything, or is it going to be split into a chain of
archive documents? If it's a chain, surely you wouldn't give each
document in the chain it's own ID, right?
The reason to use the same ID in both cases is to show that they both
represent the same feed -- that the one is the archive of the other.
