On Wednesday, December 8, 2004, at 10:21 AM, Graham wrote:
Composite feeds may have to strip them, or we could come up with a simple rule based on feed ids and headinentry.Information like this that is valid only within the context of the feed brings up issues that we need to be careful to resolve correctly. A few options and related issues:
1) Composite feeds strip this element out:
a) loses information
b) breaks signing of the entries (Could the aggregator sign the modified entry in a way that said "when I got it, I verified it's signature"--then, if you trust the aggregator, you accept that it was originally properly signed? ...what if an aggregator gets it from another aggregator which you DON'T trust? The entry would need to carry info about the whole chain of aggregators who'd signed it to enable consumers to make their own trust decisions.)
c) if this element gets defined in an extension, we'll need a must-understand mechanism to ensure that aggregators that don't know to strip it out don't pass it through
2) Composite feeds leave this in as-is, and it applies within the context of the feed whose head appears in the entry:
a) we'd either need a way to add another element specifying the priority/importance/whatever of the entry within the composite feed, or composite feed publishers would not be able to add that editorial information
3) Composite feeds add something to this element to indicate that it applies to the feed whose head appears in the entry, and may add another element to specify the priority/importance/whatever of the entry within the composite feed:
a) same as 1b above
b) 1c =~ s/strip it out/alter it/
