[Ex-plain] bib1 vs OID

Mike Taylor mike at tecc.co.uk
Fri Apr 19 12:33:13 CEST 2002


> Date: Fri, 19 Apr 2002 19:12:49 +1000
> From: Alan Kent <ajk at mds.rmit.edu.au>
> 
> Just got around to reading the attribute set definition for
> searching Explain-- records and noticed the 'record syntax - you can
> search for the OID, but its not guaranteed to find them all because
> it could be using the symbolic name USMARC'. The same was true for
> attribute set.

Oops -- that's out of date, sorry.

> Is this a good reason to say 'don't introduce symbolic names - just
> put up with OIDs so there is only one representation (put comments
> into the XML file if you want to make it clearer).

We did discuss this earlier, and the conclusion we reached (which
isn't to say we shouldn't revisit it) was that we should use symbolic
names wherever they are defined in MA's object register at
	http://lcweb.loc.gov/z3950/agency/defns/oids.html
(i.e. for pretty much everything) but with the escape route that
dotted tuples may be used for objects which are not yet registered --
such as the hypothetical Mike-Taylor-1 attribute set, which might get
the OID 1.2.840.10003.3.1000.169.42

I very much like the fact that records can therefore say "BIB1"
instead of "1.2.840.10003.3.1".  But absolutely, yes, we need to be
clear that this is _mandated_, not optional.  Otherwise we'll run into
search/data mismatches.

> It would make querying simplier if there was always a single form
> for the data (so you can do a single query instead of having lots of
> OR terms all the time).

Yes, that would be a disaster.

 _/|_	 _______________________________________________________________
/o ) \/  Mike Taylor   <mike at miketaylor.org.uk>   www.miketaylor.org.uk
)_v__/\  "If your gal don't respond when you flatter 'er, tell her
	 what Tony told Cleopaterer" -- Cole Porter, "Brush up your
	 Shakespeare"





More information about the Ex-plain mailing list