<<< OPEN LETTER FROM RAYMOND NEFF >>>
Raymond Neff
Vice President for Information Services
Case Western Reserve University
This summary has been prepared in response to inquiries concerning the
Cleveland Freenet's efforts to serve the needs of its steadily increasing
number of users. It is not intended as an exhaustive or even complete history
of the Cleveland Free-Net.
Six Facts
The Cleveland Free-Net, developed at CWRU, is the pioneering initiative in
free public computing.
Use of the Cleveland Free-Net has grown rapidly: in 1994 there were
6 million user sessions.
CWRU has invested more than $300,000 in grants and other revenue to support
and improve the Cleveland Free-Net, and continues to spend more than
$50,000 annually to operate the system.
The Cleveland Free-Net has never charged a fee to a user, and it hopes
never to do so.
The Cleveland Free-Net welcomes the arrival of commercial services, which
have begun to provide features similar to those available to Cleveland
Free-Net users for years.
The Cleveland Free-Net will remain an alternative to commercial services,
thus supporting open communication in our communities.
Background
The Cleveland Free-Net was born in 1986 as a research project in the Department
of Family Medicine at Case Western Reserve University's School of Medicine.
Assistant Professor Thomas M. Grundner was the principal investigator on the
project. As is standard operating procedure with university-based research
projects, he had the responsibility to seek funding for his research. Dr.
Grundner wrote proposals and eventually the University received funds and
computing equipment from a variety of sources to assemble the hardware and
software to create the Cleveland Free-Net. This fund raising effort raised
$44,000 from Ohio Bell, and computing equipment worth $60,000 from AT&T.
The Cleveland Free-Net system had its operational debut in July 1986. By summer
1988, the AT&T hardware was overloaded with use (it had an absolute limit of
ten simultaneous users, but it regularly ran with a limit of nine). By this
time, the system had 1,000 registered users.
During 1987-88, plans were implemented to create a more functional version of
the Cleveland Free-Net software, to be dubbed Version 2. This project was
undertaken outside of CWRU by the newly created Society for Public Access
Computing (SoPAC). Dr. Grundner was the organization's first executive
director. SoPAC engaged a computer programmer and paid $10,000 to accomplish
this objective, using funds from a grant to SoPAC for this purpose. By summer
1989, no useful software had been developed, and the project was formally
abandoned.
Because CWRU attached great importance to the evolution of the Cleveland
Free-Net, it was decided that the University's newly created Office of
Information Services should be the home base for the Cleveland Free-Net. The
University Vice President for Information Services, Dr. Raymond K. Neff, took
on the overall responsibility for the operation and development of the
Cleveland Free-Net. As of July 1, 1989, Dr. Grundner and his staff of one
full-time support person became employees of CWRU's Office of Information
Services. The University took on the following objectives for improving the
system:
Stabilizing the operation of the Cleveland Free-Net (during this period,
the computer hardware and modems had not had regular maintenance for over a
year and were unreliable).
Purchasing a computer which had the capacity to handle sixty simultaneous
users (the AT&T computer was overloaded).
Adding more incoming telephone lines.
Creating the needed Version 2 Cleveland Free-Net software.
Putting a full complement of computer professionals into the operation
and maintenance of the system (providing operational support to the system
and its users 24 hours per day, seven days per week, 365 days per year -
which is the same basis as the University's mainframe computers).
Adding more services to the Cleveland Free-Net, most notably access to
the Internet and its vast array of information resources.
By any measure, this program of improvements was a dramatic and substantial
commitment on the part of Case Western Reserve University to the still
experimental concept of free public computing.
Accomplishments
On August 16, 1989, the CWRU-developed version of the Cleveland Free-Net,
Version 2 went into public use. The six major objectives had been substantially
accomplished. The new Version 2, now registered by CWRU under the trademark
FreePort, had much more functionality than was called for in the design for the
earlier SoPAC-sponsored Version 2. This new system was based on new IBM
hardware which was far more powerful and reliable and had much more capacity
than the older AT&T system. Twenty-four modems were added to the system,
bringing the total capacity in terms of number of simultaneous dial-in users
to 34.
Usage statistics took off. Within a few months, we had added 10,000 registered
users. That early pace has hardly slackened: as of this report, the number of
registered users exceeds 160,000. The operational usage statistics show
absolutely stunning growth factors when usage for calendar year 1994 is
compared to 1988: from 160 to 16,000 individual users on any given day; 36,000
to nearly 6 million user sessions per year (averaging 28 minutes each); and
sustained growth rates of over 10% per month!
FreePort was the first version of the Cleveland Free-Net software to be based
on a scalable architecture for hardware and software, whereas Cleveland
Free-Net Version 1 had definite limitations on growth. By its very design,
FreePort could grow to handle thousands of simultaneous users, as might be
envisioned in providing "commercial-grade" public access computing. In fact,
the original target for FreePort was 600 simultaneous users, clearly in another
ballpark from Cleveland Free-Net Version 1. Today, the Cleveland system
actually supports 406 simultaneous users, including connections through all
types of Internet service providers and 260 dial-in telephone lines.
The FreePort software and the know-how associated with this novel form of
computing were soon offered to other communities who wanted to emulate the
success of the Cleveland Free-Net. CWRU has helped over a dozen other cities
start clones of the Cleveland Free-Net. A grant of the FreePort software and an
IBM computer were given by CWRU to Medina County to start the Medina Free-Net.
This missionary work looked so promising in early 1990 that Dr. Grundner
started the National Public Telecomputing Network (NPTN) as a project within
the Office for Information Services at CWRU, and after two years left CWRU to
devote full time to this concept. (NPTN has always been a separately
incorporated, not-for-profit entity.) To date, over 70 clones of the
Version 2 Cleveland Free-Net have been started.
Funding the Cleveland Free-Net Service
To anyone looking at the Cleveland Free-Net from a financial point-of-view, the
service has seemed somewhat magical. How can one offer millions of hours of
computer access time per year to thousands of people without charging them?
We have said for years that the service is free to its users, but there are
substantial costs in creating and operating the Cleveland Free-Net. Yes, the
Cleveland Free-Net is absolute free of charges to its end users, but the annual
costs of operating the Cleveland Free-Net are in the neighborhood of $50,000.
Without any external source of support for the system, these costs have fallen
on Case Western Reserve University. It is still true that, during the its
entire nine-year life, the Cleveland Free-Net has never charged a single user
for its service.
One of the possible funding models for the Cleveland Free-Net is public
broadcasting, i.e., we could use the fund-raising approach used very
successfully by PBS stations nationwide. After studying what is involved in
using this model, the University found that it did not have the in-house
expertise to carry out this approach. Over the years a few Cleveland Free-Net
users have said that they would volunteer to help raise funds for the Cleveland
Free-Net, but there have never been enough people to make the effort feasible
Another possible funding model is the one used by public libraries. Our
excellent public library systems in Cleveland and Cuyahoga County do not charge
users for the general services they provide. They are funded in part with
a portion of the property taxes levied in the region, and they also receive
a share of a State of Ohio appropriation. Public access computing as practiced
by the Cleveland Free-Net is similar to the libraries' practice of making
general information resources available without charge, and it is interesting
that the Cleveland Free-Net model has been widely accepted by the community of
librarians. Perhaps the method that funds these libraries could be adapted to
cover the Cleveland Free-Net. On the other hand, it should be noted that there
have been several levies on area ballots in the past nine years to support
public libraries, and it is not clear that the voters would be in favor of
additional tax payments to support public access computing.
Over the past five years, the University has raised some $60,000, however, by
licensing the FreePort software to other communities to start clones of the
Cleveland Free-Net. All of these funds have been used to add more hardware to
the system, especially modems. Even so, growth in usage continues to outstrip
our best efforts.
One of the ways to offer free services is to control expenses. We have polished
our skills in this respect over the past seven years. In particular, the
hardware manufacturers have continued to give the University deep discounts on
their products, and this is very much appreciated. Even so, the University has
had to spend over $150,000 for computer hardware to implement three generations
of system upgrades, including the present Cleveland Free-Net service, which
involves 18 interconnected microcomputers (16 Intel Pentiums and 486s and two
Sun SparcServers), all running UNIX. Several thoughtful Cleveland Free-Net
users have made donations of used modems to the University, and these were put
into service. As one might expect with such equipment, it did not have a long
life, and the University eventually had to buy hundreds of replacement modems.
When we did make these purchases, we were careful to acquire "smart" modems
which provided superior reliability, very high speed (from 14400 baud to 38400
baud), and "intelligent" management. These modems were relatively expensive,
but they have been true workhorses.
From our usage statistics, we see that the Cleveland Free-Net modems are in use
over 99% of all available time (over 700 hours per month each). It is no
exaggeration to say that we could use double or even triple the number we
presently have. Lack of funding has been the principal limitation to continued
expansion of the Cleveland Free-Net.
It is remarkable that the Cleveland Free-Net service now costs the University
under one penny per user session. Use of the public telephone system to connect
to the Cleveland Free-Net involves costs which add up to about four cents per
user session. Thus the total cost per user session when a telephone and modem
are used is now about five cents.
Connecting to the Cleveland Free-Net by way of the Internet is another
alternative, however. We started offering this approach in 1989, and now over
50% of all usage comes through this path. The cost of this linkage is less than
one tenth of a penny per session, so the lesson is clear: connecting to the
Cleveland Free-Net via the Internet is the most cost-effective method.
It is worth spending a moment to put these cost figures into perspective. If
the average Cleveland Free-Net user has two sessions per day via an Internet
connection, then our cost is $0.022. If the user does this every day for
a month, then the cost is $0.66 per month, the cost of two first class postage
stamps. On an annual basis, the cost is under $8.00. Given the costs to prepare
and send out invoices, we could not afford to bill our users for this service
Looking Ahead
The Cleveland Free-Net represents a pioneering effort to provide free public
computing for this community and others nationally and internationally.
Commercial services, with access to risk capital and to fee income from large
numbers of corporate and individual users, are beginning to offer many of the
features that have been available to Free-Net users for years. The Cleveland
Free-Net does not plan to compete with these commercial services, but rather to
continue to offer no-fee public computing as an alternative for those who
cannot or choose not to become customers of commercial services, as well as for
users who choose to gain access to Free-Net through connections provided by
other services. In this respect, the Cleveland Free-Net preserves the
distinctly American tradition of open communication in a new era of information
technology.
CWRU Office of Information Services
June 16, 1995
=> x=Return to Main Menu