IP: Re: Two from Edupage -- INTERNET 2'S KILLER APPS

[David Farber <farber () cis upenn edu>]

I agree .., Dave


Date: Tue, 5 Aug 1997 14:16:16 -0400
From: Lenny Foner <foner () media mit edu>
To: farber () cis upenn edu
Subject: IP: Two from Edupage -- INTERNET 2'S KILLER APPS and
  PROGRESSIVE NETWORKS AND MCI TO OFFER MULTICAST VIDEO
Cc: foner () media mit edu


    Date: Tue, 05 Aug 1997 14:06:09 -0400
    From: David Farber <farber () cis upenn edu>


[Emphasis below is mine.]


    INTERNET 2'S KILLER APPS
    Some of the applications being proposed by Internet 2 participants include
    "virtual laboratories," where researchers in geographically remote
locations
    can don goggles and data gloves to work together with colleagues using
    centralized lab equipment, and "tele-immersion," where researchers and
    students at different universities put on headsets to enter a shared
    workspace for product or architectural design.  Advanced digital libraries
> > > > could track patrons' interests via "user profiles" kept on CENTRALIZED
> > > > COMPUTERS that then automatically e-mailed digital versions of new
books or
    articles that matched a profile.  Music scholars could "jam" with
musicians
    around the country.  But one of the most difficult applications, says the
    director of the project's applications group, will be the development of a
    traffic-regulating system to provide "quality of service" -- a mechanism
    that will allow a time-sensitive transmission, such as video from an
    electron microscope, to be given priority over e-mail.  (Chronicle of
Higher
    Education 8 Aug 97)


I sure hope not!


I would be very happy to hear that this is journalistic mangling of a
more-technical story, and not that the researchers (whoever they are)
think that the right thing is centralized, searchable records of what
people take out from the library.  Real physical libraries know that
this info is quite personal and generally take pains to destroy it
when practical (e.g., not keeping infinite records of what books have
been borrowed once they are returned, including on their own backup
tapes, etc).  The major threat here is fishing expeditions and
subpoenas (and having to hassle with supporting the above), not
crackers.


There is little reason to have to have a centralized server to do
either the matching or the recordkeeping, and one hopes these
anonymous researchers actually -do- know this.


I am aware that most systems these days actually -do- have centralized
user profiles (e.g., Firefly et al), but most don't deal with the same
sorts of information being proposed above.  I'm sure the ALA would be
all over this if it started getting popular...