wasm-demo/demo/ermis-f/imap-protocol/cur/1600095084.22850.mbox:2,S

80 lines
3.1 KiB
Plaintext

MBOX-Line: From davidr at imap.cc Thu Nov 6 09:10:29 2008
To: imap-protocol@u.washington.edu
From: David Rauschenbach <davidr@imap.cc>
Date: Fri Jun 8 12:34:42 2018
Subject: [Imap-protocol] Untagged Status Notifications Async or Not?
References: <49132293.10733.4B89FB7@mail.sabahattin-gucukoglu.com>
Message-ID: <001c01c94032$923af690$0401a8c0@SYNCHRONICA.COM>
You could send an unsolicited response whenever you want, but if the client
is not in a blocking call in order to receive what you send as soon as you
send it, you might block for a very long time, in your send attempt.
Most clients are coded to read a response only after sending a request. So,
whatever you try and send from the server would not go anywhere. Therefore,
you're better off queueing your events, and flushing the queue along with
the response to the next command.
David
----- Original Message -----
From: "Sabahattin Gucukoglu" <mail@sabahattin-gucukoglu.com>
To: <imap-protocol@u.washington.edu>
Sent: Thursday, November 06, 2008 10:00 AM
Subject: [Imap-protocol] Untagged Status Notifications Async or Not?
> -----BEGIN PGP SIGNED MESSAGE-----
> Hash: SHA1
>
> Hi,
>
> I'm afraid I'm having problems with RFC 3501. The problem is that, while
> it clearly indicates that all clients absolutely MUST interpret untagged
> responses sent at any time, it seems (in a rather roundabout fashion) to
> imply that it is somehow proper for SELECTed mailboxes only ever to inform
> clients when responding to client commands, like NOOP. There's never a
> mention of why this is so, nor any permission explicitly given (as it
> seems to me) to do the right thing and just blast untagged notifications
> at the client as soon as we have them.
>
> Apart from when dealing with client bugs, I'd like to see an IMAP server I
> use push notifications at once, to have clients listen and act on them
> accordingly without requiring a minimum polling period during which they
> assume complete silence for fear of soliciting any response, and all
> without needing IDLE which we will use when we clearly need to (I.E.,
> we've quite finished interfacing with the user and are now doing nothing
> but wait for new mail).
>
> Can I be certain that it is absolutely okay to send untagged responses
> whenever we feel like it? I'd love to see this written about somewhere in
> a net note so the world will do the right thing. As things are, a few
> clients seem happy to reproduce the wait-NOOP-read-wait-NOOP cycle, which
> is totally uncool.
>
> Cheers,
> Sabahattin
>
> - --
> Sabahattin Gucukoglu <mail<at>sabahattin<dash>gucukoglu<dot>com>
> Address harvesters, snag this: feedme@yamta.org
> Phone: +44 20 88008915
> Mobile: +44 7986 053399
> http://sabahattin-gucukoglu.com/
>
>
> -----BEGIN PGP SIGNATURE-----
> Version: PGP 8
> Comment: QDPGP - http://community.wow.net/grt/qdpgp.html
>
> iQA/AwUBSRMilCNEOmEWtR2TEQKPTACgnYffKjrq8TvFQHJ8KB46eaKhc90AoOB6
> 7fpbqRwLrEO+4mIooOwoqJiZ
> =5Gr1
> -----END PGP SIGNATURE-----
> _______________________________________________
> Imap-protocol mailing list
> Imap-protocol@u.washington.edu
> http://mailman2.u.washington.edu/mailman/listinfo/imap-protocol
>