On 28/Jan/10 11:46, Ian Eiloart wrote:
--On 27 January 2010 17:22:00 +0000 John Levine <johnl(_at_)taugh(_dot_)com>
All they need to do is somehow send the message, or a pointer to a stored
copy of the message, back to the user's own mail system. The tricky bit
is that MUAs have separate configs for inbound and outbound mail, so you
can't just send the message via the outbound channel or you may be
it to someone who's never seen it before. For IMAP you could either
have a special folder (what AOL does with its IMAP interface) or a new
flag, for POP you'd have to invent a crock of some sort.
POP isn't really susceptible to simple solutions, as you say, so let's
focus on IMAP.
I really would leave this bit out for the time being. Avoiding to
upload the message is a minor optimization that can be carried out any
time. (Compare it with saving sent mail by BCC rather than uploading
the same message twice.)
IMHO it is more important to focus on _how_ and _where_, the answers
being, respectively, ARF and abuse@<authserv-id>, where the latter is
defined in http://tools.ietf.org/html/rfc5451#section-2.3 Note that
users have to configure their clients specifying which authserv-id
they trust for each account. This stone can kill both birds.
There seem to be four options, to me:
(a) define a new flag per message,
(b) define an ANNOTATE annotation per message,
(c) define an METADATA annotation on a mailbox,
(d) do something adhoc with mailboxes as is done for "spam" , "sent
messages", "trash" and the like.
Are there any other possibilities?
The IMAP UID is used to refer to a specific message (rfc5092), which
is what could be used with BURL (rfc4468) to optimize reporting when
the IMAP server can also be used as an MSA.
Asrg mailing list