{deactivating cloaking device}

I actually think that the distinction catalogers make between a
bibliographic database and an authority database is spurious.  Authority
information exists to validate the forms of name used in a catalog and to
provide a reference structure.  Only a cataloger will have any interest in
seeing authority "records" per se  and I'm not convinced we should be
building artificial distinctions into the underlying data model to keep
catalogers happy.  So I don't think we should be defining "*-authority" and
"*-catalog" indexes.  Next thing you know we'll be defining "*-holdings"
indexes too.  Neither should be be defining "*-authority" databases.

I actually think that what gets returned by a scan should be something with
a close family resemblance to an authority record.  If the term is an
unused form, it should contain information that allows navigation to or
direct searching of the used form; if there are related terms, these should
be available; if term is a node in a hierarchical structure, it should be
possible to navigate that structure.  AND it should provide an indication
of how many documents (or, preferably works) a search on the term is likely
to retrieve.

j.
RLG

{reactivating cloaking device}





                    Janifer Gatenby
                    <[email protected]       To:     [log in to unmask]
                    LCPICA.NL>                cc:
                    Sent by: "Z39.50          Subject:     Re: Scan
                    Next-Generation
                    Initiative"
                    <[log in to unmask]>


                    01/22/2003 12:37 AM
                    Please respond to
                    "Z39.50
                    Next-Generation
                    Initiative"






On second thoughts, I think it's better to use the index name to
incorporate
the target database.  So if you ask for index "author authority" it scans
that index, the occurrence count means the number of authority occurrences
and it leads to authority records.  "Author catalogue" scans give the
bibliographic occurrence count and leads to bibliographic records.  By
'loading' the index name it actually removes the ambiguity from the
occurrence count element.

On step size, I've never used it but I think it is designed primarily for
subject browse.  There are better ways to do an expanded and collapsed
scan,
e.g. by browsing headings that have no subdivisions then allowing them to
be
"opened" (Windows explorer metaphor).  Step size is a rather crude way to
do
a "quick flip".  If servers were to support some kind of expansion and
collapse, I'm not certain that many would support any step size proposed by
a client.  If we were to contemplate something equivalent to step size, I
would propose a collapsed element and leave the server to interpret its
meaning.


Janifer

-----Original Message-----
From: Robert Sanderson [mailto:[log in to unmask]]
Sent: Tuesday, 21 January 2003 20:07
To: [log in to unmask]
Subject: Re: Scan


> I'm happy with these suggestions.  I'd like to dump stepSize and
> targetDatabase.

Jannifer wanted targetDatabase. No one wants stepSize, as far as I can
tell.

Following on from my 'discovery' that index/term needs to be a
searchClause, everything fell to pieces fairly neatly:
  http://srw.o-r-g.org:8080/l5r/?operation=scan&clause=ninja

Accepts clause, numberOfEntries and responsePosition for SRU, and also
xClause for SRW. Adds in echoedRequest ala searchRetrieveResponse for SRU.
Also has a diagnostics list, but everything is Diagnostic 2 for the
moment. [Temporary error :)]

Why 'numberOfEntries' and 'responsePosition'?  These are the terms used in
ZOOM's scan, so I just copy them from the request onto the ZOOM object.
Also means for less confusion in having to do translations from one
object's attribute name to another.

Still working on the XSL for the display, but it does a simple linked
table so far which is enough to demonstrate that it works.
(Notably it doesn't use echoedRequest to generate the links, so the
searches are always against the default index and it doesn't generate
next/previous listing links)

Things which would be useful which Theo (IIRC) brought up:  Some sort of
marker for where the term would be if it's not present, or that the term
is the one used from the request.  Probably another field in <term> ?

Rob

--
      ,'/:.          Rob Sanderson ([log in to unmask])
    ,'-/::::.        http://www.o-r-g.org/~azaroth/
  ,'--/::(@)::.      Special Collections and Archives, extension 3142
,'---/::::::::::.    Twin Cathedrals:  telnet: liverpool.o-r-g.org 7777
____/:::::::::::::.              WWW:  http://liverpool.o-r-g.org:8000/
I L L U M I N A T I