* Philip Hazel:
>> This message is misleading because it is often triggered by a timeout
>> on the client side.
>
> Eh? What exactly do you mean? Do you mean that the client tries to
> connect, fails to receive the greeting in a certain time, and then sends
> QUIT? I think that is broken behaviour.
*ahem* Apparently, Exim does it, at least during callout verification.
>> Therefore, the message should be special-cased
>> for the QUIT command (or the next input should be included, like it is
>> for other synchronization errors).
>>
>> This was discovered by Debian folks, see the following bug report:
>>
>> <http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=289458>
>
> Sorry, I don't see how that report relates to what you have described
> above. That report is talking about verifying the sender address. The
> start of a connection is long before it receives a sender address. What
> have I misunderstood?
costa.debian.org times out during sender verification and sends the
QUIT message to the other Exim MTA, which logs the error message.
> I agree, however, that it would be nice if the error message included
> the next input, as it does for other errors. This may or may not be easy
> to add. I've noted the point.
Thanks.
