different explain-lite comment

  I realize that we are not reviewing explain-lite (must admit i am seeing
it more and more as a private agreement, of which I have no issue with).
But it does raise an interesting issue in terms of attribute searches
and probably distributed searches.  Actually it is a step in my implementation
direction, where everything is out of control and semanic interoperability
is probably hopeless -).


  I notice in the example there are 3 different databases, each with its own
set of search points with attributes to search - all called "bib-1".  So
there are 3 different "Institution" search points, with slightly different
attribute values allowed (yes with lots of overlap).  So it is not defining
"bib-1" (which is what you would do in EXPLAIN) - it is defining a set of
values to send to do the search when user selects "Institution".  Now
other implementations would probably chose whole other sets of values for
when the user choses "Institution".
  THink this is going the path where what you have is for a given server
the search point is:
	databaseName/searchname

so why have numeric USE values?  And why have attribute sets? 
I have been debating this for myself - once you have EXPLAIN
and the server tells the client what attributes go with 
the user selected search (e.g. "title") and these are sent back - well then
why bother? Just tell the client the allowed searchname/field names that
are searchable. And have what the user selected sent back! But I never bring
this issue up - violates all sorts of things - but since
I see the same issues appearing in explain-lite I felt it was time to point
this out.

  Bob   wald@lucent.com

Received on Wednesday, 22 November 2000 11:30:34 UTC