From: maryam@MIT.EDU
Message-Id: <9512062000.AA22349@CASPIAN.MIT.EDU>
To: acmg@MIT.EDU
Cc: acs@MIT.EDU, css@MIT.EDU, dcns@MIT.EDU
Subject: ACMG Minutes, 12/5/95
In-Reply-To: 
Date: Wed, 06 Dec 1995 15:00:12 EST


ACMG Minutes, 12/5/95

Agenda: Personal computers for undergraduates
 Guest: Joel Moses, Provost

At the outset, ACMG members introduced themselves and each proposed a
topic of interest for discussion with regards to personal computers
for undergraduates.  Following are those topics:

        - Impact on dorm start-up
        - Next 18-36 month changes affecting Academic Computing
        - Faculty assume computing power for teaching
        - Support for students using a variety of machines
        - Consistency of software
        - Availability of machines between classrooms
        - Preferred platforms
        - Joel Moses' vision of classrooms of the future
        - Obsolescence => damping? [a senior's machine will be stronger than 
                a freshman's machine]
        - New things students might be doing
        - Services to diverse machines
        - How to keep them running with minimum cost?
        - Joel Moses' vision of Athena cluster of the future
        - Architecture for the middle layer computer  
        - Evaluation
        - What about the social & human side of the clusters?
        - How are they going to deliver?
        - Servers & services

JM: Personal computers for undergraduates should not cost more than
what we're spending on Academic Computing (presently about 6.5
million) plus roughly another 3 million in the departments.  Ten
million dollars is a lot of money and it's worth asking whether we
could do more, if not right now, then in 5 years.  But asking the right
questions is very important.  For example, one solution is a
three-tier computing power:

        1. What would be a cheap device that would provide basic services
           (perhaps e-mail, and word-processing, not graphics)?
        2. A middle ground computer
        3. A powerful computer


DISCUSSION:

It can be argued that the higher level is clearly our responsibility,
and that the lower level is a student's responsibility.  Who is
responsible for the middle ground?

Problems with technology transfer are readily apparent.  Even within
MIT, technology transfer is not done right for several reasons.
Research universities like MIT must work with industries on technology
transfer.

File service and common file space will still be a problem.  Students
with a basement level laptop cannot now easily access their home
directories.  This should remain transparent. While it is true that
the WWW is a top layer that ignores the various machines' operating
systems, it has done so at the expense of write access.  Efficiency
and scaling at the basement level is most important.  Because we pay a
lot of attention to issues of scale, we spend much less on staff to
support our environment than do other comparable schools.

One thing we need to do in order to see whether this three-tier
model would fit our needs is to gather statistics on what software
people are using in the clusters.  A simple way to accomplish this
might just be to go out there and interview people.

Joel Moses was asked what does he want to accomplish with this.  His
answer was to make sure we are doing things the right way.  The answer
may be yes.  The big change being in communications, it is worth
asking these questions even if changes are to happen not immediately
but in 5 years.  However, asking the right questions is very
important.  Distance learning and video-conferencing are changing the
face of teaching methods.  And soon the split between computing and
communications will not be as clear as it is today.

There is a task force being formed by Michael Dertouzos, Director of
Lab for Computer Science, with a one-year mission; its membership
has yet to be finalised.  Its purpose is not to re-examine Athena but
to examine all on-campus computing activities part of which is
Academic Computing.  Another longer term group led by William
Mitchell, Dean of the School of Architecture & Planning, called the
Council on Technology in Education, will examine both on-campus and
off-campus potential uses of technology.


	Next meeting: December 19, in room E40-212


----------------------
Maryam Afshar
Academic Computing Services
MIT, Room E40-359
Email:  maryam@mit.edu
Phone:  253-1346