If the an SMTP Server allows mail from unspecified sender to a valid user, it would be an open relay. So Mail with header MAIL FROM: <> would be rejected to guard against spam. Is this a requirement to satisfy SMTP AUTH Extension ? I could not tell by reading some of the docs.
Now getting on to the practical problem If a mail is sent to an invalid external user, the external server will typically catch the mail and generate a bounced message. The external server does this to notify the sender that the mail could not be delivered. The problem is that typically the Mail Server sending the generated bounce message does not specify MAIL FROM. If this command 'MAIL FROM: <>' is rejected by the server, effectively bounce messages are not received. What should be the correct behavior ? A good mail server should guard against spam as well as receive bounced message notifications. How could these be reconciled ? Harmeet -- To unsubscribe, e-mail: <mailto:[EMAIL PROTECTED]> For additional commands, e-mail: <mailto:[EMAIL PROTECTED]>
