Yes, that accords with my understanding. (And I'll take another look at that README.) Thank you.
On Fri, Apr 21, 2017 at 9:39 AM, Bill Cole < postfixlists-070...@billmail.scconsult.com> wrote: > On 20 Apr 2017, at 17:54, J. Johnson wrote: > > Incidentally, I do have "smtp_delay_reject = yes", which *delays* the >> processing. But I appears that none of the smtpd_*_restrictions (if >> specified) are skipped. Which I find handy for controlling how they are >> applied to different addresses. >> > > Correct. See the SMTPD_ACCESS_README document and the postconf(5) man page > for details of how smtp_delay_reject operates. It's primary purpose is to > avoid problems with how some SMTP clients react to rejection to the > commands that come before RCPT. It is important to understand that while > smtpd_delay_reject=yes gives earlier restriction lists > (smtpd_client_restrictions, smtpd_helo_restrictions, and > smtpd_sender_restrictions) access to restrictions that use inputs not > available until after their logical phase, you still get in-order > evaluation of the restriction lists and of the restrictions within each > list. It is still a good idea to NOT use late inputs (e.g. recipient > address) in early restriction lists unless you have a specific need to do > so. >
