On Fri, 25 Apr 2008, Ken Keating wrote:
For testing purposes we built 2007a to accept secure connections on port 994
rather than 993.
How did you do this? Did you change IMAPSSLPORT in imap4r1.c from 993 to
994 (the most straightforward way)? Or did you use some other means? If
so, what?
Outlook, Thunderbird, and Mac mail on Leopard (OS X 10.5)
seem to work just fine, but Mac mail version 2.1.3 on Tiger (OS X 10.4) has
problems. We get error messages in our sendmail log:
Apr 18 09:45:31 <hostname> simapsd[10190]: [ID 960700 mail.info] Unexpected client
disconnect, while reading line user=??? host=<fqhn> [<IP addr>]
This isn't an error message. It simply means that a not-logged-in client
disconnected the session without first issuing a LOGOUT command. Some
clients do this routinely in the notion that this is "more efficient".
So, all we really know here is that Mail.app on Tiger did not log in, and
then disconnected the session.
I can think of two likely possibilities off the top of my head:
[1] You are mistaken, and your port 994 server is not really doing SSL.
The client expects to do SSL, communications never start, and eventually
it disconnects.
[2] Your port 994 server is doing SSL, but the client is not expecting
SSL, communications never start, and eventually it disconnects.
To dismiss [1], do
telnet imapserversystem 994
You should be rewarded by dead silence after the connection is
established. If you see an IMAP greeting banner, then [1] is the problem.
To verify that [1] is not the problem, run openssl and do
s_client -connect imapserversystem:994
You should see a bunch of SSL diagnostic crud, and eventually an IMAP
greeting banner.
To test for [2], run a packet sniffer and try a connection. If both
sessions are quiet after connecting, that suggests [2], since a plaintext
IMAP client expects to be spoken to before it speaks, and in SSL the
client starts speaking first.
Finally, turn on debugging telemetry in Mail.app. I don't know how to do
that, but I know that it can be done. See what the log there says. If
you see any IMAP protocol negotiations, either figure them out yourself or
pass it on to me and I'll tell you what it was doing.
-- Mark --
http://staff.washington.edu/mrc
Science does not emerge from voting, party politics, or public debate.
Si vis pacem, para bellum.
_______________________________________________
Imap-uw mailing list
[email protected]
https://mailman1.u.washington.edu/mailman/listinfo/imap-uw
