Philip Hazel <ph10@???> probably said:
> files. Without wild cards, if the incoming local part matches some
> alias, then that's good enough verification. I can see that with wild
> cards there might be a case for doing more, but the problem then is that
> an alias might expand into several different addresses, and what do you
> do if some of them are OK and others are not? I suppose the most
> straightforward thing is to insist that all of them verify.
Hrrrm.
If at least one of them verified then the message could be delivered
to some of the recipients - I would say that is a configuration error,
it should deliver to those it can and stick (freeze ?) on the queue
for the rest.
I'd like to see recursive verification through wildcards and IMO only
fail if none of them are valid eventually.
P.
--
pir pir@??? pir@??? pir@???
