The Economist’s year-end double issue carries an intriguing story about how consumer Internet services are slicing into areas traditionally controlled by IT (information technology) departments. GMail, Skype, Yahoo! Instant Messenger, and other services developed for consumer use are facilitated by companies that use massive data centers, “built to fail” server software, and redundant network connections. A few IT managers are starting to recognize that companies like Google succeed or fail based on the uptime of their services. The old maxim, “We need to keep it in house because we’ll be able to keep it safer that way” doesn’t hold as much weight these days. To Google, Skype, Yahoo!, et. al., time offline is lost profit.

According to the article, IT manager Adrian Sannier is using a pilot program from Google called Google Apps for Your Domain, which allows managers to use any combination of Gmail, Google Talk, Google Calendar, Google Page Creator, and a Start Page, all under a private label. I was happily surprised that Mr. Sannier works for Arizona State University. My experience with university IT departments over the years has been that they are usually home to understaffed, overworked, underpaid employees who are often stretched in many directions. Unfortunately, many of them are also managed as fiefdoms locked in the eternal struggle for budget dollars. Ceding control to end users is a scary thing to many IT departments. Centralized management of hardware and software is in some instances a necessary evil, but in many cases it becomes the raison d’etre of the IT organization.

Information technology departments are not profit centers. In a university, the development department obtains grants and other funding. The office of admissions decides how to price tuition. Academic departments and the marketing folks try to lure the right kind of professors, who in turn ostensibly attract the most desirable students. The way some people in universities see it, the IT department just spends money. Every time some new technology comes along, IT hears the same refrain: “Hey, that looks easy. Why don’t you whip something up that will do that for us?” It makes sense that IT folks should seek silver bullet software packages that enterprise vendors call “solutions,” because they’re always being asked to do the impossible.

But software silver bullets don’t often work as advertised. One of the great things about consumer-focused technology like GMail is that in order to be successful, it has to be easy enough for regular humans to install and use. Enterprise applications like GroupWise, on the other hand, don’t operate like they’re designed for regular humans. They’re designed for their customers, who are of course IT departments. That may explain why such apps try to do everything. The unfortunate result is that they seldom do anything very well. In particular, apps that were build around the old centralized model of corporate computing then adapted years later to the Internet, tend to function like ducks on dry land. They get along, but without much grace.

Enter Google applications. They’ve been built not as a whole, but as individual apps that live on the Internet and function either in isolation or together. If you use GMail, you don’t need to use Google Calendar. But if you do, there’s some nifty integration between the two. Seem familiar? This is in some ways the online equivalent of what Apple has done with many of its apps. Mail, iCal, and Address Book integrate quite nicely, but you can use any one of them independently, and substitute one of them with another product without penalty. Mail won’t penalize me if I use Google Calendar rather than iCal to plan my life.

The fact that Google and Apple took similar approaches may have to do with the UNIX philosophy. One of the key tenets of UNIX is that “small, sharp tools” work best. This philosophy is part of the reason why there are so many different flavors of UNIX (FreeBSD, OpenBSD, NetBSD, Linux, Solaris, etc.) and so many different small applications that do the same thing in different ways.

Collaboration and communication via the Internet is becoming more powerful and less expensive every month. Many people already use web mail, a web-based calendar application, web-based bookmarking, web-based RSS feeds, and web-based collaboration tools. These are essentially commodity applications. Why then, should university students be forced to use clunky, bloated, out-of-date software that would be laughed at if it were introduced in the consumer market? Why would an IT department want to deal with the hassle of maintaining software that requires huge up-front investment and multi-year support and training contracts? Why would a university administration want to waste IT funds on commodity tools when it could be spending them on wireless networks, tech training for professors, better IT staff availability for students, and development of those custom applications that are truly necessary?