The Sticky Side of the Web



Copyright 1996, University of Colorado. From _CAUSE/EFFECT_ 
Volume 19, Number 3, Fall 1996, p. 48. This article is 
reprinted with permission from the May/June 1996 issue of the 
_Digit_, 

http://www.colorado.edu/CNS/Digit/
 
For further information contact Wendy Dubow at 
wendy.dubow@colorado.edu


              THE STICKY SIDE OF THE WEB
              by Kenneth J. Klingenstein


Of the various inventions implemented over the last twenty 
years on the Internet infrastructure, none has had the impact 
that the World Wide Web has had. Begun as a tool to assist 
scientific researchers in sharing scholarly data, the Web 
arrived at the right moment, when the fusion of computing and 
digital communication was reaching fruition. The result has 
been breathtaking growth and the beginning of a remarkable 
societal transformation. The Web, in its ease and power, 
offers a profound new mechanism for information and 
communication. Among its many benefits, it promises economic 
opportunities, new efficiencies for government and social 
services, and significant environmental relief. But not all 
of its effects have been positive. In dimensions ranging from 
network impact to cultural preservation, from laws to mores, 
the Web presents difficult challenges in an urgent time 
frame. 

From being a mote in network traffic two years ago, the Web 
is now over 50 percent of total Internet volume. Its ease of 
use, both for navigation and content presentation in 
multimedia and interactive formats, has brought legions of 
new users and new uses to the network. And in its propensity 
for complex graphics, the Web represents a massive number of 
bytes to convey. (A picture may be worth a thousand words, 
but most graphics contain an order of magnitude more bytes 
than a thousand words.) Multiply new users and uses by such a 
volume per usage, and the current severe network congestion 
is the result.

Moreover, all this traffic is carried by a somewhat 
inefficient protocol that is insensitive to network load. 
Most network applications employ a "slow-start" approach that 
transmits packets at a limited rate until the current 
carrying capacity of the network is determined. This "good 
citizenship" by applications helps deter the chronic death-
spiral that congestion can create. HTTP, the underlying Web 
transport protocol, does not employ slow-start, likely 
because its designers assumed that the typical Web user, 
flitting from host to host, would have such brief connections 
as not to justify implementing slow-start. Thus, the large, 
bursty transfers of Web data can cause, and exacerbate, 
network congestion.

If the Web is technically boorish, it is a social radical. 
Its power as the ultimate printing press means that it can 
push the limits of our societal bounds. For example, while it 
presents powerful opportunities for diverse populations to 
inform the rest of the world of their own lives, it 
represents a real threat to those in the community who are 
trying to preserve their cultures from contamination.

Many of these issues are illustrated at universities, where 
our traditional leadership in networking has brought us to 
the forefront of its associated problems. We are having to 
create new policies to deal with self-expression on the Web. 
When a student in a dorm room, operating on his or her own 
machine, posts offensive material on the Net, what authority 
does the institution have? Who determines if an image is 
libelous or harassing? Such issues were once largely 
academic, but given the reach of the Internet, these 
questions now have world-wide impact and implication. Indeed, 
the Net's global scope complicates the matter by removing the 
context of local standards for policies.

The Web also poses significant threats to our personal 
security and privacy. In particular, while Web clients help 
users to glean information from the servers, the clients also 
can convey information, surreptitiously, from the client back 
to the server. Today the clients may report to the server the 
identity of the client's owner; within months they will be 
able to report the contents of the user's hard drive and 
convey files back to a server. While legal precedents have 
been established in other areas of privacy (for example, 
barring public disclosure of what books you have checked out 
of the library or what videotapes you have rented) there is 
no guidance on privacy in cyberspace. The temptations for 
businesses, or government, to know details of the people who 
visit their Web sites may lead to unforeseen abuses.

The Web, much like other transformational inventions, has 
consequence far beyond its first assessment. While we have, 
as institutions and as a society, already begun to reap the 
benefits of the Web, we have yet to deal with its sticky 
side.


*************************************************************

Kenneth J. Klingenstein (ken.klingenstein@colorado.edu) is 
Director of Computing and Network Services at the University 
of Colorado, Boulder, and a member of the CAUSE Board of 
Directors.