On Wed, 10 Dec 2003, Arnt Gulbrandsen wrote:
> I don't think so. In my opinion, no commercial software vendor should
> pretend to support something without testing it. Is that callous? Am I
> being hard on the poor overworked programmers of the world?
If you support IMAP4rev1 according to RFC 3501, then by definition you
support IMAP2.
An IMAP2 client never asks for UIDs. Therefore, a compliant server never
sends UIDs to an IMAP2 client.
Let's consider what would happen if it were otherwise:
There would be no reason for the UID message data item. All the text
related to it can be deleted.
Furthermore, clients would have to be written to accept any future message
data item, even if no such thing is specified. So, in order to be
prepared for what IMAP69 servers might do, all clients must have the
following secret (not in RFC 3501) rules in their clients in order to
avoid being broken in the future:
msg-att-dynamic =/ msg-att-future
msg-att-static =/ msg-att-future
msg-att-future = atom SP future-value
future-value = future-value / future-list / future-other
future-list = "(" future-value *(SP future-value) ")"
future-token = atom / ("\" atom) / nstring
future-other = <new syntax dreamed up by our successors>
; client authors MUST have ESP and recognize ALL
; possibilities
I don't think so............ :-)
-- Mark --
http://staff.washington.edu/mrc
Science does not emerge from voting, party politics, or public debate.
Si vis pacem, para bellum.
