Re: [exim-dev] questions on BMI code

John Jetmore wrote:

> This entire email can be summed up with "Is there a reason the BMI code
> runs after local_scan rather than before it". Details below if they
> matter.

My reason for this decision was that local_scan is the last point where
you can reject after data, and BMI scanning is a very costly operation,
so it should be the very last thing to do in the chain. You can look
into receive.c, it should not be very hard to change the order.

I'm also using BMI in "dumb" mode just as you do (I have a simple
blackhole router that just nukes the BMI spam for everyone). Brightmails
whole concept does not fit in well with rejects after data (It likes to
keep spam in a quarantine, reinjects messages, has explicit
multiuser/domain support etc.). But I guess you know all this, it's just
an excuse for the list. :)

/tom