On Wed, 9 Oct 2002, Alan Kent wrote:
> On Tue, Oct 08, 2002 at 01:13:37PM +0100, Robert Sanderson wrote:
> > How can we say which combinations of relation modifier and relation are
> > supported by individual indexes?
> I agree the matrix is ugly to implement. I was going to take the cop
> out of just letting the server reject the query. Otherwise you have
> to list all the relations etc per index as well - not all indexes
> might support adjacency or greater than. It seems too verbose to
> list all the relations etc per index name (it would work and would
> be accurate, it just would be ugly).
Yep.
People won't do it if it's too hard, tedious or just plain useless.
I would think that the following rules would be useful:
* If the document specifies a setting for an index, this takes precedence
over any contradictory setting in the global configInfo.
* If the document specifies a setting in the global configInfo, then this
applies to all indexes, not otherwise overridden.
The only situation this leaves as tedious/hard/stupid is if one index
doesn't support something and all the others do, but I'm prepared to say
that this is a valid situation for just letting the server fail the query.
Rob
--
,'/:. Rob Sanderson ([log in to unmask])
,'-/::::. http://www.o-r-g.org/~azaroth/
,'--/::(@)::. Special Collections and Archives, extension 3142
,'---/::::::::::. Twin Cathedrals: telnet: liverpool.o-r-g.org 7777
____/:::::::::::::. WWW: http://liverpool.o-r-g.org:8000/
I L L U M I N A T I
|