[LibX] LibX question from a faculty member

Ross Singer ross.singer at library.gatech.edu
Mon Nov 27 19:45:19 PST 2006


On 11/27/06, Godmar Back <godmar at gmail.com> wrote:
> Abigail,
>
> I suppose the question is whether these interfaces are publically
> documented or easy to reverse engineer.

>
> In the "SFX prechecks Aleph" case, the interface is probably Z39.50.

Well, in the SFX/Aleph case, SFX has an 'unfair advantage' in that
(obviously) Ex Libris has the ability to query the SQL tables
directly.  That allows it to be a lot faster and more efficient than
Z39.50.  With our link resolver we do roughly the same thing (except
we export our bib db to IndexData's Zebra and for holdings coverage we
query Voyager directly through Oracle).  It's /much/ faster than a
direct Z39.50 query to Voyager.  It requires a lot of inside knowledge
about a particular ILS, however.

> In the "Scopus checks SFX" case, I'd be curious to know what interface
> is being used. Do you have more information about it?

SFX has the ability to resolve multiple openurls at once.  Scopus
sends all of the context objects on the page to SFX and displays an
icon according to what it gets back.  And, yes, it's my understanding
it uses AJAX (we don't subscribe to Scopus, but I thought I remembered
Rolf Kwakkelaar of Elsevier mentioning something about this in a
presentation).

Dave Walker of Cal State San Marcos takes a much more pragmatic
approach (in my opinion) with his dynamic SFX icons in Xerxes (his
metasearch engine).  He parses the Google Scholar export file for
holdings coverage which is much faster than querying SFX.

-Ross.


More information about the Libx mailing list