Dick St.Peters wrote:
In 5.3.6(a) alias forwarding, the return code is 250, not
251.
Without checking, 1123 didn't chage the return codes. The
interestig case for SPF is "user not local", other aliases
within the same organization (no 3rd party MX involved)
have no potential problem with SPF.
251 is for telling the sender that the user isn't local
but you'll forward. I believe the intent was a hint to
the sender that he/she should send mail directly if
possible. A typical use would be where a user had moved
and changed address.
Yes. Either 251 or 551. 251 is a temporary courtesy,
it makes no sense to waste bandwidth with more points of
potential failures if the sender can send it directly.
From a 821-POV forwarding to 3rd parties was a temporary
courtesy / hack, not a permanent solution.
A common use is a user who is temporarily relocated
Yes, then you use 250 - but note that it's still meant
to be a _temporary_ solution, not _permanent_
Another is a user who receives mail at different
addresses for different purposes or businesses
and wants to deal with it all from a single mail
account.
Many ISPs offer to collect mails from various POP3
accounts (not exactly better than forwarding... ;-)
The reason why I say 251 is the "user not local" bit,
because that's the critical case for SPF. Not the
complete 5.3.6(a), only if an unrelated 3rd party MX
is involved (that has not white listed the forwarder
and checks SPF)
I'm not saying that 251 is "better" than 250 in this
case, there's a long chapter in 2821 why 251 / 551
might be a bad idea (as in your vacation example).
But the effect of SPF FAIL caused by 251-forwarding
is like 551, if neither forwarder nor next hop did
anything about it. A simple solution is 5.3.6(b) -
behave like a mailing list.
Bye, Frank