Mailing-List: contact cygwin-help AT cygwin DOT com; run by ezmlm List-Subscribe: List-Archive: List-Post: List-Help: , Sender: cygwin-owner AT cygwin DOT com Mail-Followup-To: cygwin AT cygwin DOT com Delivered-To: mailing list cygwin AT cygwin DOT com Message-Id: <3.0.5.32.20050523195343.00b6bc08@incoming.verizon.net> X-Sender: phumblet AT incoming DOT verizon DOT net Date: Mon, 23 May 2005 19:53:43 -0400 To: "Gary R. Van Sickle" , From: "Pierre A. Humblet" Subject: Re: I need some exim hand-holding In-Reply-To: <0IGV00AFYH6DJCXB@vms046.mailsrvcs.net> Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" At 10:46 PM 5/21/2005 -0500, Gary R. Van Sickle wrote: >Ok, I give. No mas, no mas. > >I have a new mutt build ready to go, but I can no longer test it since I've >gone DSL. AT&T does [E]SMTP over SSL, and ssmtp (which I had been using) >doesn't (the official Cygwin one anyway). So I'm trying to set up exim as a >sendmail replacement, and of course getting millimeters away from it >actually working and then being stopped cold by unknowns. > >I'm going to try building ssmtp with SSL support, but in the meantime, can >somebody tell me what the secret is to getting exim to work in this >situation: > >- This is on XP, NTFS, ntsec on. >- This is send-only, as mutt has POP built in. >- I'm trying to use it in this manner, in the Cyg-specific README's >*nontsec* section: >"6) Mail User Agent programs can use exim to send local and remote > mail. If you are using mutt, add the following in .muttrc > set sendmail="/bin/exim -i" . To use exim from cron, point the > symlink /usr/sbin/sendmail to /bin/exim ." >- I'm using mua_wrapper, and not using the daemon at all, so does this >section still apply? >"8) Due to file permission settings (there is no suid in Windows), > Mail User Agent programs cannot call exim directly for local > deliveries in a secure environment. They should smtp to exim on > localhost (port 25). > This restriction can be lifted (while maintaining some security) > by using the -odq option and giving append permissions to the > spool, input and msglog directories and the log files, and read > permission to the /etc/exim.conf file. > Deliveries are then made by the daemon on the next queue run. > This restriction is also lifted if all modes are 0777 or 0666." >I'm getting no failures that look like file permission problems. >- I appear to have TLS/SSL set up correctly, since exim gets to the point >where it receives a 221 and sends an EHLO or HELO. But then the AT&T server >(apparently) just sits there, and after a few minutes drops the socket. >- I've read everything I can find and FWICT what I have in /etc/exim.conf >should work. I have of course tried many other permutations of options as >well. Attached if a kind soul would care to look it over. > >The results of an "exim -d g DOT r DOT vansickle AT att DOT net" are also attached. Late >breaking news: that was done using the same exim.conf which is attached, >with my ATT username and password munged, but I get the same results when I >have my real UN & PW in (what I believe to be) the proper exim.conf >locations. > >So is it a Cygwin problem, an exim problem, or a Gary problem? I'll let you >guys know how the ssmtp+SSL thing goes. From you debug output: 204.127.134.146 in hosts_avoid_esmtp? yes (matched "imailhost.worldnet.att.net") not sending EHLO (host matches hosts_avoid_esmtp) If you don't send EHLO, the functionality is reduced. Pierre -- Unsubscribe info: http://cygwin.com/ml/#unsubscribe-simple Problem reports: http://cygwin.com/problems.html Documentation: http://cygwin.com/docs.html FAQ: http://cygwin.com/faq/