On Sat, 15 Mar 2003, Florian Weimer wrote:
> Exim doesn't react gracefully to an initial 554 greeting:
It's clear from the discussion that there is disagreement over this. Let
me first give you some history:
Exim used to treat 5xx on connection the same as 4xx. Then somebody
suggested that it should be changed because, otherwise, what is the
point of having both 5xx and 4xx if they are both treated the same?
So I changed it.
Two general points:
1. Exim has always operated on just the first digit of the SMTP return
code, as the RFC suggests is possible. So I wouldn't like to make an
exception for 554 on connection if I can avoid it.
2. Saying "try the next MX" is all very well - but there may not be
another MX. Or maybe they all give 5xx on connection.
If the managers of the domain want to reject all mail, then Exim's
behaviour is what they have requested. If this is some kind of mistake,
then Exim's bouncing of mail will draw it to somebody's attention
quickly. Waiting for 5 days before bouncing is not, IMHO, so useful.
> I don't feel very strong about this issue. I need to change this code
> snippet at work, I think, but I'm not sure if I'd want to force this
> change on all Exim users.
While I think Exim now does the best thing, I would not be against
adding an option to make this switchable. It's a trivial change, and if
it makes everybody happy....
On Sat, 15 Mar 2003, Giuliano Gavazzi wrote:
> 554 Transaction failed (Or, in the case of a connection-opening
> response, "No SMTP service here")
Right. So if an address is routed to such a host, something is screwed
up (unless this is deliberate). Bouncing the message causes someone to
pay attention.
> What I am less sure about is if it is appropriate to fail all
> addresses when the 5XX error (504, 550) is in response to the HELO.
Again, I argue that if you don't, you are treating 5xx as 4xx, which
seems wrong.
On Sat, 15 Mar 2003, Florian Weimer wrote:
> Some people use the 554 greeting to indicate "no mail under this
> domain", some to say "please try the next MX".
Well, I think they should use 4xx to mean the second of those. Then
there is no ambiguity.
Philip
--
Philip Hazel University of Cambridge Computing Service,
ph10@??? Cambridge, England. Phone: +44 1223 334714.
