[Ex-plain] Protocol Version (Was: Attribute Searches)

Mike Taylor mike at tecc.co.uk
Fri Apr 12 14:54:00 CEST 2002

> Date: Fri, 12 Apr 2002 12:54:09 +0100 (BST)
> From: Robert Sanderson <azaroth at liverpool.ac.uk>
> > > The term of Protocol and the term of ProtocolVersion aren't
> > > really related in terms of categorisation... one's a name the
> > > other is a number (or similar).
> > 
> > Sure they are!  Search for "z39.50" alone and you want all z39.50
> > (nor SRW) servers; bung in a "version" SQ of "3" and you're asking
> > for version-3 Z39.50 servers.  Seems perfect to me.
> How do you know that you want *Z39.50* version three?

Because the access-point is "version" and the term is "z39.50".  The
semantic qualifier says, "Oh, and not just _any_ Z39.50 server -- I
only want version 3 servers".

> Can you give an example of how you would actually do the search?

find @attr NET 1=4 @attr NET 2=version3 z39.50

> Or are you saying we should do:
> find @attr NET 1=4 z39.50 AND @attr NET 1=4 2=1 3 


> What's the benefit of having it as the same access point with a
> semantic qualifier if you have to always AND it together with the
> same access point for it to make sense?

There would be no benefit to this.  The semantic qualifier is a single
attribute which provides additional precision on the access point.
It's not an SQ of type "version" with value "3", but an SQ of value

> This says to me that it's a second access point.

No -- who would ever want to search for version3 alone?  That is, for
all servers that support version 3 of a protocol, but we don't care
which protocol?  That's meaningless.

 _/|_	 _______________________________________________________________
/o ) \/  Mike Taylor   <mike at miketaylor.org.uk>   www.miketaylor.org.uk
)_v__/\  "(12+144+20 + 3*(4^(1/2))) / 7 + 5*11 == 9^2+0" --
	 unattributed arithmetic limerick, from /usr/games/fortune.

More information about the Ex-plain mailing list