Given a choice here the extension "x-databases=abc,def,ghi" seems the more
natural. I would prefer to get something more extensible for MSEs to be
called with, to take account of other needs they may have which target SEs
do not have (such as the need to de-dupe results, possibly using a
particular algorithm). I would like to see this become an extension which
could be used for all MSEs (at least as far as common functions are
concerned), and so would suggest parameters such as "x-mse-databases=...."
or "x-mse-dedupe=....." as a scheme that could be adopted.

It does take the SRU/W model away from its original meaning, but it seems to
me that this is an important enough area of our business (a little bit of
bias showing here, perhaps?) that we should attempt to create something to
bring the MSEs into the same fold. This leads to the fact that sections
26-29 of the Spec (to do with extensions) mention "profiles" and 'profile
designers" quite a few times with no definition of what a profile is in this
context. Is it (part of) a context set? (but they are for CQL?) If not what
is it? Confusion is reigning (or raining) at the moment.

I'm trying to get to a point here where a set of extensions could be
registered by a group of MSEs (and other interested parties) so there is a
standard extension to handle this traffic. If we can't get this done, then
we will have MSE specific extensions or else have to wrap the whole SRU/W in
another session protocol which sort of defeats the purpose of simplicity,
even if preserving a 'division of purpose' clarity. A similar issue arises
with respect to authentication, bit that can come up for discussion later.

Peter


> -----Original Message-----
> From: Z39.50 Next-Generation Initiative [mailto:[log in to unmask]]On Behalf Of
> Dr Robert Sanderson
> Sent: Monday, June 13, 2005 12:19 PM
> To: [log in to unmask]
> Subject: Re: Multiple Sources in SRU/W
>
>
> > As far as I can tell from the documentation and from searching
> this thread
> > there is no preferred/recommended/allowed method of asking a server to
> > search more than one target database. Obviously with "real" SEs
> this is not
> > a problem as they only have one database (well, the vast
> majority of them),
> > but for SRU/W to be used to connect TO an MSE it is obviously a bit of a
> > problem. From the MSE to the SEs is no problem.
>
> That is all correct.  The SRW model is that different databases are at
> different end points.  As opposed to OAI and Z39.50, where you can access
> different databases or record sets from one location.
>
> An extension 'x-databases' extension could work, or by putting the
> database names or identifiers into the URL for the service.
>
> eg:
>
>      srw.cgi?...x-databases=a b c d
>
> or:
>
>      http://foo.com/a/b/c/d/srw.cgi?...
>
> Rob
>
>        ,'/:.          Dr Robert Sanderson ([log in to unmask])
>      ,'-/::::.        http://www.csc.liv.ac.uk/~azaroth/
>    ,'--/::(@)::.      Dept. of Computer Science, Room 805
> ,'---/::::::::::.    University of Liverpool
> ____/:::::::::::::.
> I L L U M I N A T I  Cheshire3 IR System:  http://www.cheshire3.org/
>