On Fri, 12 Apr 2002, John W Baxter wrote:
> Philip, does such added testing--triggered by -bV, perhaps--seem reasonable
> as a post-book wish list item?
I'm afraid it doesn't seem very reasonable to me, because it would be so
very limited. Only unmodified names (i.e. nothing that involved a change
of name via an expansion), and only in the main part of the
configuration, and only when running as the user who called Exim (-C
discards privilege). The best it could do would be to output "xxx does
not exist" for files it couldn't find - but there are configurations
where files don't exist all the time, and Exim in fact makes use of this
fact.
I fear that this would cause more confusion that it would dispel.
There is too much logic involved in the use of various files to build a
test that isn't "the real thing". Of course, there are some better tests
than -bV, which really is just a syntax check. You can use -bt and -bh
to test quite a lot, and there is also -N.
--
Philip Hazel University of Cambridge Computing Service,
ph10@??? Cambridge, England. Phone: +44 1223 334714.
