Proposed Solutions

• Rely on existing rules set in network usage documents (for example, Stanford's Computer and Network Policy)
• Inform users that all actions are logged and monitored, and that bandwidth limits have been set.
• Remove functionality that infringes on the rights of the user, effectively crippling the software.

Rely on set policy (top)

       If a policy is already in place, perhaps the easiest solution is to forego the addition of packet-monitoring software and rely on the documents. For example, Stanford University's Computer and Network Usage Policy outlines acceptable use of network resources. Part of the document states that users of the network "have a responsibility not to abuse [its] resources and to respect the rights of others." Abuse of network resources includes failure to respect copyrights, which can be taken to refer to not only software but digital music and video. Since there is already a rule on the books regarding use of such, a rule that can be applied to Napster and its brethren, it might be unnecessary to purchase software to remedy the situation. The administration could interpret the rule as described and begin to crack down on students using the aforementioned services. This would inevitably lead to freed up bandwidth.

       Also, Stanford's policy covers situations where a user is sharing legal files that are large and bandwidth-consuming. Section 2-b-2, "Encroaching on Others' Access and Use," forbids the sending of excessively sized files, data, or programs when doing so could have a negative impact on the network. Section 2-e-2, "Personal Use," states that only "purely incidental" use of the network is appropriate for "personal activities not related to appropriate University functions."

       With the above three rules already in place, it seems there is an existing law concerning use of network-intensive applications. If the Computer and Network Usage Policy was enforced at Stanford, the bandwidth problem would vanish. Violators of the rules would be punished, and this would deter others from using the network in an unacceptable manner. In addition, the system administrators would not need to know exactly what/where users have been doing/going. They would notice that a user has been generating a lot of traffic, and could take appropriate action.

Inform users of changed policies (top)

      This is akin to the employer telling the employee that his/her email is being monitored. It is the right of the company to ensure its own privacy, and the worker must accept it. This is a possible solution only if it is agreed that the university/student relationship is equal to that of an employer/employee. If this is true, then the user's rights can be sacrificed at the benefit of the network. For the administrator, this is ideal ö the packet-monitoring software retains its full functionality, and the network can be shaped to work at maximum efficiency. However, the employer has to let the employees know that this policy is in effect. When Stanford University started using Packeteer's Packetshaper, students were not notified until the school newspaper, The Stanford Daily, printed an article about the technology. Newsgroup postings such as the following were common:

Newsgroups: su.computers.rcc
Subject: Re: erratic network access

No, you're not alone·. At some points in the day, the network will be very slow, but at some points, it just cuts out completely.... Our residents aren't happy at all about this, naturally. The funny thing is, they've stopped coming to me when things like this happen, because they've come to expect it.

Furthermore, even today there is no information online about Stanford's use of PacketShaper. The ITSS' "Changes to Computing Services" (ITSS is Stanford's Information Technology Systems and Services department) site does not mention the software, neither does Stanford's Residential Computing department page.

       Users must be informed of potentially serious changes to the networking system, especially those that potentially reduce their level of privacy or disrupt their use of a paid network service. At Stanford, students pay $80 a year for what they assume is uninterrupted, unlimited Internet access. They did not sign any agreement from the University stating that their access would be capped, or that the administration would have the power to choose what traffic should be let through at certain speeds. Many students are still unaware that a packet-monitoring solution is being used on the Stanford campus. If network administrators made an effort to inform their users of service and policy changes, users would be more informed, and more capable of making intelligent decisions regarding use of the network.

Remove non-essential elements of software (top)

      Another possible solution is forcing makers of packet monitoring software to remove features of their software that might infringe on the rights of the end-user. If this is infeasible, students could at least petition to their system administrators to not use the privacy-invading options. The software provided by Packeteer is very powerful, and it has the ability to help decongest the network even without personally identifying those that frequent certain sites. For example, a general bandwidth limit on peer-to-peer software will effectively relieve the media-sharing problem, and finding out specific information about users is of little additional help.