[Greylist-users] Bagley: version 0.01 available

[William Blunn]

> If the ACL implementation fixes the cases like certain Exchange servers
> and YahooGroups that do not retry (can someone confirm this please) then
> this is worth taking a look at.

It depends what the problem is.

If the remote MTA is ignoring an error after the end of the DATA segment (and erroneously assuming that the delivery was successful), then an Exim ACL implementation could help to address this problem.

However, my guess is that the problem in this case is that the MTA is interpreting the temporary failure as a permanant error. In this case it ought to send a bounce to the sender.

So we ought to be able to tell if an ACL implementation would help by seeing if those MTAs send a bounce back to the sender.

If they are sending a bounce, then chances are they will probably bounce on a RCPT temporary error as well, so an ACL implementation wouldn't be any better.

If they aren't sending a bounce, then chances are they are ignoring errors after the DATA segment, in which case an ACL implementation should be better.

I don't have any data on those MTAs.

Bill