Author: James P. Roberts Date: To: exim-users Subject: Re: [Exim] local_scan() addition: views sought
> > I think this request is/was bogus. And the feature is really not
needed > > at all. Noone is interested in the message that comes with a success
> > return code.
>
> Since I am the one who requested the "bogus" change, let me explain.
>
> If I return a 250 but toss the message to /dev/null, this is very evil. > The least I can do is return a message saying so to help debugging if the > message ever has to be traced.
> It also helps when you talk to the smtp server yourself for debugging.
>
> Marc
Just a minor comment, which y'all may feel free to ignore:
If you have accepted a message, and choose to bit-bucket it because you
are sufficiently certain it is spam (or virus, whatever), then I would
declare it is NOT evil to fail to mention this to the sender. (No more
evil than deleting the message from your inbox, based on the subject
line, for example).
On the other hand, the debugging point appears to be quite valid.
Therefore, I would suggest the possibility of a second API, something we
might call local_scan_debug(), for lack of any better imagination. Or
perhaps add something to the source that can be compiled in optionally?
This need not need be coded by Philip, I should point out. Hurrah for
Open Source!
