Proposed "IDLE" extension comment

Mark Keasling <makr@winter.airco.co.jp> Mon, 10 March 1997 06:43 UTC

Received: from cnri by ietf.org id aa29947; 10 Mar 97 1:43 EST
Received: from mx1.cac.washington.edu by CNRI.Reston.VA.US id aa02125; 10 Mar 97 1:43 EST
Received: (from daemon@localhost) by mx1.cac.washington.edu (8.8.4+UW96.12/8.8.4+UW96.12) id VAA27434 for imap-out; Sun, 9 Mar 1997 21:42:40 -0800
Errors-To: owner-imap@cac.washington.edu
Sender: owner-imap@cac.washington.edu
Received: from ns.airco.co.jp (ns.airco.co.jp [203.178.5.68]) by mx1.cac.washington.edu (8.8.4+UW96.12/8.8.4+UW96.12) with SMTP id VAA27427 for <imap@cac.washington.edu>; Sun, 9 Mar 1997 21:42:36 -0800
Received: from airco.co.jp (ns.airco.co.jp [203.178.5.68]) by ns.airco.co.jp (8.6.9/3.3W97022115) with ESMTP id OAA04610 for <imap@cac.washington.edu>; Mon, 10 Mar 1997 14:44:24 +0900
Received: from winter ([192.168.10.66]) by airco.co.jp (8.6.9+2.4Wb3/3.3W97022115) with SMTP id OAA26935; Mon, 10 Mar 1997 14:30:58 +0900
Message-Id: <199703100530.OAA26935@airco.co.jp>
Date: Mon, 10 Mar 1997 14:42:51 +0900
From: Mark Keasling <makr@winter.airco.co.jp>
Reply-To: Mark Keasling <makr@winter.airco.co.jp>
Subject: Proposed "IDLE" extension comment
To: imap@cac.washington.edu
In-Reply-To: <SIMEON.9703061027.E@mars.watson.ibm.com>
X-Mailer: AIR MAIL for Motif (v1.5)
MIME-Version: 1.0
Content-Type: multipart/mixed; boundary="MimeMultipartBoundary"

Hi Barry,

Can a server send additional responses between the time the client issues
DONE and the OK response is returned if a final check for expunges, exists,
alerts or whatever doesn't take 10 minutes (ie. can be done reasonably
quickly)?  The draft implies that this can't happen with the statement:

   The IDLE command is terminated by the receipt of a "DONE" continuation 
   from the client; such response satisfies the server's continuation 
-> request.  At that point, the server MUST immediately send the tagged 
-> response to the IDLE command and prepare to process other commands.  

and therefore MUST NOT do anything else.

On Thu, 6 Mar 1997 10:55:27 -0500, Barry Leiba <leiba@Watson.IBM.COM> wrote...
> Please take a look at
> 
>    ftp://ietf.org/internet-drafts/draft-leiba-imap-idle-00.txt
> 

Looks good.  Slightly different than what I saw discussed where IDLE
was terminated by the next command.  However, it solves a possible race
condition where the server sends an expunge at the same time a client
issues a FETCH command.

Mark Keasling