On 18/5/05 7:41 PM, "Graham" <[EMAIL PROTECTED]> wrote: >> Some are content changes, or metadata changes. > > I see no discussion of this distinction in the atom:modified proposal.
I meant "content or metadata changes". I don't care so much about the distinction between those two senses. I was however making a distinction between those and representation/format changes. The atom:modified proposal refers to changes in "the entry", not changes in "atom:entry". Subtle, I know. > atom:updated does not have to be user selectable. It's perfectly > valid to leave it as the first publishing date, or to use a last > modification date. > > (Not that either of these are useful and one is actively harmful, in the sense of being polluting. >I think atom:updated should be optional, but Tim doesn't listen) the problem is that if atom:updated is optional we then have to spec the fall-back precedence for when it is missing, and also spec what it means if an entry had atom:updated, but is then published without atom:updated. can of worms. e.
