On Oct 15, 2005, at 1:32 PM, Roy T. Fielding wrote:
Given that the format ties the semantic of this element to what
the "publisher considers significant", it seems logical to
conclude that the protocol needs to have a way to indicate what
the publisher considers. Whether this is done with a
pub:control field or by explicitly setting the value of updated
(i.e., whether the atom:updated values are sourced from server or
client) is worth debating.
It should be content within the atom entry being published.
The way that is defined makes it no different from a title.
I am definitely in favor of accomplishing this using something within
the entry that you PUT rather than extraneously via a header, because
it means that the message is more completely self-describing.
Consider the case where I want to do an "insignificant" update. I
see three options:
1. remember the last time I updated, and include the same value for
<atom:updated>
2. have an explicit signal as in PaceInsignificantUpdate
3. leave atom:updated out of the body you PUT, then document in the
protocol spec that this has the semantic that the client does not
consider this significant.
#2 or #3 both seem sane. I prefer explicit signaling and thus lean
to #2. -Tim
