I'm unhappy with the opinion that not specifying a schema might result in
records from multiple schemas being returned. Servers should be expected to
specify their default schema (through explain) and return all records in
that schema when an explicit schema has not been specified.
> -----Original Message-----
> From: Ray Denenberg [mailto:[log in to unmask]]
> Sent: Wednesday, July 10, 2002 10:05 AM
> To: [log in to unmask]
> Subject: Re: sort parameter
> Mike Taylor wrote:
> > > Date: Tue, 9 Jul 2002 13:49:52 -0400
> > > From: "LeVan,Ralph" <[log in to unmask]>
> > >
> > > Well, the model is that the records aren't in any
> particular schema
> > > and they get put into the requested schema on the fly.
> So, when you
> > > ask for them to be sorted, you ask that first the records be
> > > transformed to a particular schema and then you sort them
> on a field
> > > (or fields) from that schema. Doing this, there is no mixture of
> > > MODS and DC records. There are only records in the requested
> > > schema.
> > Huh. This is new to me. Matthew, is that what you'd
> understood too?
> > Your example of surrogate diagnostics seems to imply not.
> There are a few issues raised by this exchange.
> You're mixing the sort/resultSet model with the present model.
> Addressing the present model (leaving aside sorting for the
> moment) there is
> the question of whether in a sequence of returned records can
> two records be
> presented in different schemas. Obviously so if you include
> the surrogate
> diagnostic schema, but can one record be in MODS and another
> in DC? Yes, if
> the schema parameter is omitted and the first record is
> available in MODS
> only and the other in DC only. Suppose the record schema parameter is
> included, and it's MODS? We haven't answered this but I
> suspect the answer
> is you get a surrogate diagnostic for the one that's not
> available in MODS.
> Finally, the sort model you quote no longer applies, we
> discarded it several
> weeks ago. We're not sorting result sets according to a
> schema, but rather
> by abstract names.