Postfix reject_unverified_recipient question

[Peter Ong]

----- Original Message -----

> From: "Glenn Steen" <glenn.steen at gmail.com>

> And did you verify this, empirically? I don't think you did, because
> (according to that last resort: the docs;-) Postfix should be smarter
> than that.... Read the section about "reject_unverified_recipient"
> again. My interpretation is that the unverified_recipient_reject_code
> only affect things that generate a 5xx return code, so the
> "soft_bounce default behaviour" is only there so that it doesn't mess
> things up while you test the recipient verifiacion out (dependong on
> whether you have control of the detsination, it being correctly
> configured for rejecting unknown recipients etc, this is prudence).

Yes, I verified it quite extensively.

(http://www.postfix.org/ADDRESS_VERIFICATION_README.html)
The unverified_recipient_reject_code was confusing in this regard when I read it at first. But upon closer examination in Postfix 2.6 there is an unverified_recipient_defer_code which clued me into maybe that "reject" was supposed to be modified. At first, I feared that it was going to unconditionally reply 5xx to anything even for lesser offenses. But now that I see the "defer" version, I'm quite convinced it's okay to change it to 550.

After the change, I've tested it and it works. However, I am unable to create a test situation where the final destination server would reply in the 4xx or less. But I think it's working now. Thanks Glenn and everyone.

p