On Fri, May 03, 2002 at 03:41:49PM +1000, Matthew Hawkins wrote:
> On Fri, 03 May 2002, Craig Sanders wrote:
>
> > On Fri, May 03, 2002 at 02:53:47PM +1000, Matthew Hawkins wrote:
> > > header and body checks are done by cleanup, not smtpd, so the
> > > advantage is already defeated [*].
> >
> > why say things you know are untrue?
>
> Quoting cleanup(8)

you know damn well what i meant. stop wasting everyone's time with
pathetic attempts to score "points" over pedantic crap.

in case it really is beyond your comprehension, you said:

a) "...so the advantage is already defeated [*]."

then later:

b) "[*] just to confuse people, the advantage is not lost at all."

you knew the advantage wasn't defeated at all - not in the slightest -
so why say it was?

 
> (Just to name 3 references to back up my claims that header/body checks
> are done by cleanup, and that cleanup is called by smtpd)

it should be damn obvious by now that that is NOT what i was querying
you about. i was querying your disingenuous statements about when the
reject takes place and what affect that has on whether a bounce is
queued or not.

what i don't get is why you got so defensive. you made a suggestion, i
pointed out some flaws in it and suggested a better way. what's there
to get so defensive about?

> > your suggestion was to receive the mail without body/header checks
> > then, depending upon recipient address, re-inject it back into
> > postfix with header/body checks enabled.
>
> My suggestion was simply to use this feature of snapshot 20020331:

and as i said, i pointed out that doing this would probably have
undesirable consequences. if you accept a message and then test whether
it is allowed by your policies then you have a duty to deliver a bounce
report. if you test it against your policies before accepting it and
reject it outright then the SENDING system has the duty to deliver the
bounce.

avoiding unneccesary work for your server (especially pointless work
caused by undeliverable spammer addresses) is always a Good Thing. it
makes little difference if you only have a handful of undeliverable
bounces in your queue at any given moment, but it makes a huge
difference to performance if you have hundreds or thousands of
them....they're all taking up resources, cpu power, network bandwidth,
syslog time & disk space, etc that could be used for real mail.

the way to avoid getting hundreds or thousands of them is to reject
undesirable mail with 5xx codes rather than accept & bounce it.

craig

-- 
craig sanders <castaz.net.au>
Fabricati Diem, PVNC.
 -- motto of the Ankh-Morpork City Watch
-
To unsubscribe, send mail to majordomopostfix.org with content
(not subject): unsubscribe postfix-users

• Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]