[Mimedefang] Accessing the source/destination port #'s at filter_relay

[Philip A. Prindeville]

On 05/04/2010 01:15 PM, David F. Skoll wrote:
> We could generate the ID right away, as soon as the other end connects,
> so it would be available from filter_relay.
>   

How likely is that to happen?

We've gone production here with my patches... to eat our dogfood... but
if it's likely that the reworking of MdF happens such that
read_commands_file() is now available from each of the filter_xxx()
hooks, that's equally good and we could go with that too if such a patch
is to be forthcoming that we can try out.

Just trying to figure out what decision has been made:  adopt (possibly)
my patch into the source tree?  Or patch MdF to use a different
command-file naming scheme that would allow earlier access.

I've no clue how much work is involved in the latter, hence no idea how
likely it is to come to fruition any time soon.