On Wed, 30 Jul 2003, Dean Brooks wrote:
> Now, I'd really love a way to specify that if the transport_filter fails
> to continue on with the transport instead of deferring, but that's
> probably asking a bit too much. ;)
Continuing on is probably a bad idea given that one doesn't know what
state the delivery is in (parts of the message might have already passed
through the filter and been written to the real output, for instance).
However, given that the delivery will be deferred, you could adjust the
routing so that it didn't pass through the filter on the 2nd or
subsequent delivery attempt. However, there is no guarantee that the
deferment was caused by a filter failure, so this is probably not a good
idea. I can think of various nasty ways of kludging this up, the tidiest
of which is probably to put a wrapper round the filter that that takes a
decision as to whether to run the real filter or a dummy pass-through.
--
Philip Hazel University of Cambridge Computing Service,
ph10@??? Cambridge, England. Phone: +44 1223 334714.
Get the Exim 4 book: http://www.uit.co.uk/exim-book
