Johan Almqvist wrote:
(and maybe we should modify qpsmtpd's behaviour to be more consistent in terms of what it returns, so as to issue 5xx's for the rest of the transaction if a DENY has been returned).
Beat you to it. There is code in branches/0.3x now that disconnects any connection where we've already issued a 5xx once (except for QUIT and RSET).
John
