> with "suffix = -approval" the -approval address works until it gets expanded to the owner- address and then
> majordomo2 is skipped. here exim is functioning correctly but maybe exim should check the address before
> the pre/suffix is stripped when determining if the address has been previously directed.
The current action assumes that the director is making no use of the
prefix/suffix when handling the address; consequently, when checking for
loop prevention, it is only the stripped local part that should be
considered. [Originally, variables such as $local_part_prefix weren't
provided.] I see your point about checking the prefix/suffix as well. I
guess it won't do any harm, because even if a director isn't using the
prefix or suffix, it would still only go round twice in a looping
context, for example if "prefix-user" got turned into "user", running it
through again would produce "user" and so it wouldn't run it through a
third time.
> the docs say that the prefix and sffix options are string lists but with multiple suffixes, none are matched. here
> is the debug output with "suffix = -outgoing:-request" and "no_suffix_optional":
That is a bug that was fixed in Exim 1.62, released on April 17th.
> and finally, i get sementation fault on the -outgoing address. here is the debug output with "suffix = -outgoing":
> directing morty@???
> Segmentation fault
Exim 1.61 has the following bug fix, which may well be this one:
2. Exim could crash when handling multiple directions for a single message,
especially if caused by several chained aliases.
Regards,
Philip
--
Philip Hazel University Computing Service,
ph10@??? New Museums Site, Cambridge CB2 3QG,
P.Hazel@??? England. Phone: +44 1223 334714
