reprinted from the Penn Printout:

BRINGING THE NETWORK HOME

Al Filreis

I have begun to write this essay in Van Pelt College House, in the Faculty Master's apartment, at 10:30 PM on a Tuesday night near mid-semester. At the moment thirty-seven residents of this small House, and two non-residents, are logged on. Four residents and the two non-residents are working in the computer room (our "Reslab"); the other thirty-three, including two of the resident Faculty Fellows and four of the five Van Pelt Graduate Fellows, are using computers attached to high-speed ethernet connections ("ResNet") in their private rooms. By the time I have drafted this, at midnight, the number of Van Peltians working on line will be greater, not smaller. The numbers will begin to drop at 1 or 2 AM.

This is when students - and faculty, as you can see - do much of their work, and it is when they do their computing. And so it should be when - and where - primary computer support ought to be available.

In some ways this is already the case. During a recent informal meeting with President Rodin held in the House, a dozen Van Peltians were asked where they get help with computing. Eleven answered: right here at home. When students here in the House have a question or problem, they turn to members of their own residential community. They go to the computer room, for instance, where during most of the hours of the seven days a "computer consultant" - a work-study student hired within the House - is on duty. Or they turn to members of the new Van Pelt "Sandbox Committee", a group of info tech-savvy volunteers from within the House who have been set free to design a workable, flexible "primary support" system. The response can be astonishingly fast. One evening in mid-December, two Van Pelt residents, Angela and Alexis, found to their horror that, in their not very precise phrase, "someone had erased Netscape" from the computer they used in their room. When they attempted to download Netscape again, they suddenly lost their electronic mail interface, "Host Presenter." They found Elliott, a computer consultant, who in turn posted a message to the Sandbox Committee listserv. Minutes later, one Sandbox volunteer, Becca, who remembered that a similar problem had occurred elsewhere in the House, and who has emerged as our specialist in installing particular kinds of ethernet cards in the Resnet system, responded. Becca happened to be in her room, and made a house call, checked the ethernet card manual and setup program, and did another install. Angela and Alexis had had term papers to write and were awaiting crucial course-related e-mail. A problem that in the best of circumstances would have taken several days to address, and in any event not by a house call, was solved within an hour. And who knows but that left to their own devices, unaware of the right way to turn, they might have done further damage.

The Sandbox solution described above wasn't just free (though that is surely a significant factor), and didn't merely save a full-time computing staff person a good deal of time at an already busy time of the academic year. What's perhaps more important is that it just makes good intellectual sense at a residential university. It empowers students to help students, and to see themselves formally in the business of teaching and advising in an area where they have largely untapped expertise. It strengthens the self-sufficiency of a collegiate community.

Let us invest in our collegiate communities by setting up primary support systems on the principle that the best support is that provided with local (and personal) knowledge. Those involved in the restructuring of computing services across Penn are articulating just this sort of principle. They contend that ideally "[t]he primary support provider is known and readily accessible to the end user. In many cases, the primary provider is physically co-located with his/her users." For undergraduates this can and should mean the delivery of this vital service through the emerging collegiate system. For the students on campus, one imagines the Sandbox model - and that of the very successful Science and Tech Wing of King's Court/English House - being formalized and replicated across the residential system. The many students who do not live on campus, since they will affiliated with a collegiate unit, will seek support day or night in that unit - with its hub or locus for delivering this and other student services.

People in computing call this a "distributed environment" - localized and locally responsive. Penn is at its best, I think, when academic programs and services in general are "distributed." Many non-academic services are more efficiently centralized than spread and replicated across the units. Once computing was among these. It should be no longer. Innovations are local, but require local support. Support should be awaiting the needs of these innovators - and perhaps even to some extent should anticipate them. (The "primary care" analogy works well when one considers the advantages here of "preventive care." If Becca had originally been the one to install Angela's and Alexis's card, the problem in December might not have arisen.) Computing enables the academic life to be increasingly greater than the sum of its parts, but only if the parts are cared for. Newly productive intellectual combinations, enhanced by powerful information technology, are beginning to gather as effectively in the "loft spaces" of the networked residential system by night - in the Reslabs, in students' rooms, in common lounge spaces with their ethernet boxes - as in the classrooms by day.