> > Because if you implement callbacks, you're most likely not refusing <>
> > What I already discussed with Philip was to have callbacks that first do
> > MAIL FROM: <>
> > fail null mail from
> > or
> > RCPT TO: <dest>
> > fail null mail from
> > RSET
> > MAIL FROM: <postmater@mysite>
> > RCPT TO: <dest>
> > -------------------------
>
> What happens if dest does a callback? How do you avoid "callbackloops"?
I don't think there would be a loop if you did this properly.
Ie:
A: sender somerandom@???
B: contact whereever.com
mail from:<> fails
mail from:<postmaster@???> works
rcpt to:somerandom@???
C: notices the connection and does a callout for postmaster@???
At this point mail from:<> should not fail.
postmaster@??? should be accepted w/o using sender verification
B: says it's ok
C: says it's ok
B: accepts mail from A
A = the sender
B = the MTA the sender is sending to
C = the MTA that holds the sender's email account.
I'm assuming in this case that B also holds the postmaster@???
address.
--
Lab tests show that use of micro$oft causes cancer in lab animals
