Cataloging the Internet

[Peter Graham, Rutgers University Libraries]

From: Peter Graham, Rutgers University Libraries

>Books go out of print.  URL's become unavailable. Etc.  Sure it happens at a
faster rate but what's the diff?<

This fatuous comment touches on a point others have made, including whether
libraries (librarians)_ should catalog the whole Web/net or not.  It seems to
me likely that when we look back in the future we will see a high correlation
between cataloged data objects and data objects whose stability and integrity
is assured in some fashion, most likely by their being held by organizations
we call libraries which will assure their continued existence in usable (and
integral) form.  

Stu Weibel has commented on some of the ancillary tools to make this
possible:  persistent naming structures, for example.  Others include
institutional commitments to preserve data over time, authentication
technologies, refreshment and technology migration strategies, and disk
farms.  

Those who point out that we don't catalog everything now are on the right
track.  There's nothing automatic about librarianship, either in print or
electornically:  we make value decisions every day.  An example that
demonstrates the rule is the existence of various (printed) ephemera
collections, e.g. the John Johnson at Bodley and the one at Bowling Green U. 
I'm sure there will be some servers in 2195 spinning old Web4Lib files, home
pages of high schools of 1997, and a few grainy UU-encoded porn shots just as
examples of what used to be the bee's knees.  --pg

Peter Graham    psgraham at gandalf.rutgers.edu    Rutgers University Libraries
169 College Ave., New Brunswick, NJ 08903   (908)445-5908; fax (908)445-5888
              <URL:http://aultnis.rutgers.edu/pghome.html>