>>>>> "Marc" == Marc G Fournier <[EMAIL PROTECTED]> writes:
Marc> I've thought to modify the code, to extend the spam
Marc> extension, so that it adds a simple:
Marc> X-Spam-Score: <True/False> <score> / <threshold>
Marc> so that if spam is enabled, then it auto-adds this header,
I thought about doing something like this, but decided that actions
that modify messages as a side effect would be a bad thing.
Marc> fillin_header() in sieve/script.c looks good, and is after
Marc> the spam checks are run/scored, but am not 100% certain of
Marc> how I should call add_header() for the above ...
fillin_header() is used to build notify responses -- it doesn't do
anything to the incoming message.
I think the right place to do this is either in an lmtp proxy or
inside lmtpd before it calls sieve. I'm playing with code in lmtpd
that allows the administrator to specify a zero or more filter modules
that are called in sequence where sieve is called now. Sieve would
become just another filter module. This would allow the system
administrator to set up filtering policies that would be difficult or
impractical to do within sieve alone.
-- Bob
