Date:         Fri, 7 Jan 2005 14:59:42 -0000
Reply-To:     "Z39.50 Next-Generation Initiative" <[log in to unmask]>
Sender:       "Z39.50 Next-Generation Initiative" <[log in to unmask]>
From:         "Matthew J. Dovey" <[log in to unmask]>
Subject:      Re: Adlib Base profile
Comments: To: "Z39.50 Next-Generation Initiative" <[log in to unmask]>,
          [log in to unmask]
Content-Type: text/plain; charset="us-ascii"

 > > Tolkien
> > 
> > is equivalent to some query of the form
> > 
> > index1 Tolkien or index2 Tolkien or index2 Tolkien or ...
> 
> That is specifically what I *don't* do.  I don't want to OR 
> together 100 terms for every unqualified term entered.

I meant that the result set is the same (rather than the actual
mechanism for doing the search).
 
> > My semantics for serverChoice and serverDefault are that they are 
> > equivalent to
> > 
> > index Tolkien
> 
> That is exactly what I use serverChoice for.
> 
> > Where that index is a single index which could be 
> explicitly used in a 
> > CQL query. From Ralph's description, you can't do an equivalent
> explicit
> > CQL query of the form
> > 
> > BasicIndex Tolkien
> 
> But I can do cql.serverchoice Tolkien

Yes - but my interpretation of cql.serverchoice Tolkien is that there is
some non wildcard CQL index available such that I can issue both

cql.serverchoice Tolkien 

Or

index Tolkien

Against the same server and get the same result set back.

In you case your omitted index, and serverchoice semantics are closer to
the anyField or anyKnown in that there is no other equivalent cql query.

Actually having written than, why isn't what you are doing an anyKnown
search (in that BasicIndex is an index not accessible by CQL?)

Matthew