shadypalm88
08-06-2003, 06:39 PM
I am in MTA hell. Sendmail is insecure. Qmail is rebellious and won't touch anything remotely related to aliases. Niether seemed to have any kind of native SMTP authentication support. So now I'm at exim. Looks good on paper. Has authentication, and is supposed to work fine with mailman. I'm saved at last. Not quite. Well it compiled (finally). And it runs as a daemon. This is all good. After much fiddling, swearing, praying, etc. I can now e-mail local users. Hurray!
However, I'm back to square one on e-mailing people on other domains. Adding my IP address and the IPs of everyone using my system to some relay accept list is not an option. This is why I want to use SMTP authentication. The question is: how? Remote users who I want to have mailboxes and be able to send mail may or may not be actual system users. I tried to compile with MySQL checking ability but it wouldn't compile (yes, the MySQL client/libraries are installed). The plaintext auth example from the exim specification works fine, but it's just for one user. How can I get a viable multi-user checking system?
However, I'm back to square one on e-mailing people on other domains. Adding my IP address and the IPs of everyone using my system to some relay accept list is not an option. This is why I want to use SMTP authentication. The question is: how? Remote users who I want to have mailboxes and be able to send mail may or may not be actual system users. I tried to compile with MySQL checking ability but it wouldn't compile (yes, the MySQL client/libraries are installed). The plaintext auth example from the exim specification works fine, but it's just for one user. How can I get a viable multi-user checking system?