The pace adds this under atom:feed:
atom:feed elements that contain no child atom:id element
MUST contain an atom:link element with a relation of "self".That could just as well be:
atom:feed elements that contain no child atom:link element
with a relation of "self" MUST contain an atom:id element.Each subtly implies a preference for one or the other. I prefer the second--a preference for an element that can't be spoofed (or where spoofing can easily be detected). If that's missing (as in cases like SMTP delivery that have been noted, where the feed isn't retrievable from a specific URI), then atom:id must be added to provide a way to identify the feed.
I didn't change the Pace, since such a change could conceivably change the opinions of people who've expressed an opinion on it already. But I would be interested to know whether people think this would be an improvement, make it worse, or if either is just as well.
As for 'atom:feed elements SHOULD contain at least one atom:link element with a relation of "alternate".', I don't care particularly whether this is SHOULD or MAY, just not MUST. So if another Pace makes this a MAY, I as the author of PaceFeedIdOrSelf, would not consider it to conflict (on that point, at least) with PaceFeedIdOrSelf.
