Re: New ANNOTATEMORE draft
Rob Siemborski <rjs3 <at> andrew.cmu.edu>
2003-06-17 17:31:51 GMT
On Wed, 21 May 2003, Cyrus Daboo wrote:
> A New Internet-Draft is available from the on-line Internet-Drafts
> directories.
In addition to Ken's question, I note that the draft stipulates:
"Servers SHOULD allow annotations on all 'types' of mailbox, including
ones reporting \Noselect for their LIST response."
At the very least, we might need more text here (or a justification of the
SHOULD as opposed to a MAY, if we really want this at all).
Say we have:
user.rjs3
user.rjs3.foo.bar
Then, user.rjs3.foo may be returned as a \Noselect mailbox.
Say I set an annotation on user.rjs3.foo, and then delete the
user.rjs3.foo.bar mailbox. What should be the behavior of the
annotation on user.rjs3.foo?
Does the annotation go away and then come back if I recreate user.foo.bar?
Does it continue to exist independently? If either of these is true, does
this mean I SHOULD be able to set an annotation on
arbitrary.mailbox.that.may.not.exist.currently? I don't think this sort
of behavior should be encouraged.
(Continue reading)