Let’s say you are a new first year Dartmouth student. Your mom and dad purchased a Mac for you to use during your senior year in high school. You arrive at Dartmouth College, one of the fifteen or so colleges in the USA that requires every student to own a computer. You discover to your dismay that Dartmouth has a Mac standard that, in fact, permits only computers purchased from Dartmouth to be connected to the Dartmouth network. The policy does not allow you to use your home-bought computer on the network at Dartmouth. This type of standard simplifies connecting large numbers of computers to a network and makes things easier for Dartmouth’s computer support staff.
Williams College, on the other hand, accommodates a variety of computers. Student computers, for example, have many different kinds of network interface cards (NIC), each of which has its own software and peculiarities. A NIC is the device in a computer that connects it to EphNet at high speed. Supporting this degree of NIC diversity alone has direct consequences for the speed with which a computer can be connected to EphNet.
It is self-evident that the Dartmouth model is easier and cheaper to support, because it has just one platform, one operating system, and one network protocol. At Williams College, the STCs (Student Technology Consultants) must include two groups, MacForce and WinForce, to support student users of Macs and PCs, respectively. STCs support not only Macs and PCs, but also multiple varieties of platforms, operating systems, and network protocols. Working out all the possible combinations when connecting to EphNet or any network is not always trivial.
Finding a balance between rigid standards and flexible ones is a challenge. The greater the mix of problems brought to a small service group like the STCs, the more difficult it is to respond quickly. Many schools like Williams College have support policies that identify different categories of support levels. The highest level includes unqualified support for word-processing, spreadsheets, email, and so on. These types of software are typically used all the time and immediate help should be available. The middle level includes less frequently used software and gets less support. The third category includes software that is rarely used and receives little or no support or support only when time is available.
We’re working on developing simple clear standards. We presently do our best to support the following on both Macs and PCs: MS Word, MS Excel, MS PowerPoint, MS Access/Filemaker Pro, Power Term InterConnect/32, HyperTerminal, Netscape, virus software (FPROT and Macafee), and Eudora. We support Novell file services and UNIX time-sharing functions like email (PINE).
We have literally dozens of academic software programs (close to 200 in fact) that we are sifting through to determine reasonable levels of support. We have five or six different statistical programs, for example, and we have to determine what levels of support we can provide for each one.
Some standards are relatively easy to introduce.
Recently, for example, we’ve had to adopt Windows NT as a new standard for Jesup lab PCs. Because Windows 95 can be modified easily, we’ve had to introduce an operating system (NT) that can be set up to prohibit local changes. In this particular example of shifting from Windows 95 to NT, users should experience absolutely no difference in functionality or look and feel. Introducing other software standards, however, can take long campus discussions to work out.
In principle it would be nice if we could support all the software our users could possibly want. We can’t. So, our goal must be to identify clearly what we can support and to execute that support well.