On Wed, 2004-03-03 at 10:47 +0000, Nigel Metheringham wrote:
> Tests of this and details of experience using it would be quite
> useful...
See also
http://www.infradead.org/rpr.html for a limited implementation
done entirely within Exim, although this doesn't implement a local
database storing details of the addresses it generates, as is required
to fix the multiple-forwarding-with-SPF case safely.
Instead, my implementation just doesn't attempt to special-case the
multiple-forwarding case. SRS breaks for long local-parts _anyway_ so
I'm happy just to treat already-rewritten addresses (with their
lengthened localparts) as normal breakage introduced by the whole
SPF+SRS+... system.
--
dwmw2
