[Top] [All Lists]

Re: I-D ACTION:draft-ietf-sieve-notify-xmpp-00.txt

2006-05-17 14:08:05
Hash: SHA1

Dave Cridland wrote:

On Wed May 17 11:35:21 2006, Alexey Melnikov wrote:
Personally, I'd be happiest with a URI to the new message, where
possible, but i appreciate there's security concerns there.
What form would such a URI take?
Can be an IMAP (draft-ietf-lemonade-rfc2192bis-02.txt) or HTTP URL.

Could reasonably be a mailto: URL, too, which might merely identify the
email address the mail arrived at, as a fallback. 

True. Or, presumably, a pop: URI.

(OMA's EMN format uses
a non-IANA registered scheme for that kind of thing, too).

Non-IANA-registered schemes are scary to me.

IMHO the account and message identification considerations belong in
draft-ietf-sieve-notify rather than in the XMPP "profile" thereof (which
would simply provide an instantiation of those considerations).
Maybe you are right, but I don't think that draft-ietf-sieve-notify
can prescribe any particular URI type, because this would vary from
one deployment/implementation to another.

Sure. Presumably we can say that the account and message SHOULD be
identified by a URI, but that the scheme to use or prefer shall be
determined by local service policy.

It could say "Notifications SHOULD include a URL for the recipient to
use as a hint for locating the message. Where the nature of the
notification allows, this SHOULD be marked in a machine-readable manner
(such as an attribute in XML)."

I'd strike the parenthetical clause in sieve-notify and in the xmpp
profile say something about how to encapsulate such a URI in an
appropriate XMPP extension (e.g., JEP-0066).


- --
Peter Saint-Andre
Jabber Software Foundation

Version: GnuPG v1.4.1 (Darwin)
Comment: Using GnuPG with Mozilla -


Attachment: smime.p7s
Description: S/MIME Cryptographic Signature

<Prev in Thread] Current Thread [Next in Thread>