On Mon, 7 Aug 2000, Dave C. wrote:
> Hrm. I'm not sure if specifying that as part of the search-type makes
> sense, since the search types are a specification of what to search in
> and what protocol? to use.. What you are looking up seems like it
> should be seperate - I know this is sort of a step backwards, but
>
> hostip_accept_relay and hostname_accept relay
>
> seem to be the most logically way of making the distinction.
I see your point, but the problem then is: what order should the tests
be done in? At least in the current state it is clear - `first check
that the IP is this, then check that it's that, then check the name is
such-and-such' and so on. One of the problems with the historical
`backwards' state was in deciding on (and then explaining!) how multiple
options interacted with each other.
--
Philip Hazel University of Cambridge Computing Service,
ph10@??? Cambridge, England. Phone: +44 1223 334714.
