Giving Up on the Silver Bullet
My generation must have watched too many "Lone Ranger" episodes in our youth. I know I watched every week. No matter how terrifying the fight became, up galloped the Lone Ranger to save the day just in time. We now imbue technology with the same magical qualities. Week after week, we believe that the next technology solution will fix all the problems left by the last invention.
But in the public sector-and particularly in municipal government-placing faith in the next wave of technology is particularly dangerous. Usually our capital budgets are small, and our operating budgets are thin. New technologies cost a lot of money; the infrastructure almost always needs refurbishing; training demands time that has been allocated to other things. And because virtually no technology is as simple as it appears, confusion reigns following implementation.
When we become frustrated at such business conditions, we start thinking about technology as a pendulum rather than as a silver bullet. Instead of cheering on the next wave, we notice that things have started to swing back to the way they were. But if we can remember the pendulum while planning, we can incorporate what will be needed in the next wave while implementing the current one.
Planning for the "counter-trend" while concentrating on implementing a new technology can mitigate the risks of each advancement.
The client/server trend is a good example of this cycle. Some years ago, the question was: Are you going client/server? This is the ultimate example of a bad technology question. None of the "uninitiated"-that is, our customers-could penetrate the meaning of this simple question, composed entirely of common English words. The question focused completely on a particular technical solution, with absolutely no reference to business interests.
However, client/server did try to address some genuine concerns. Inflexibility had grown up around the mainframes that were then in existence. User acceptance of large systems was diminishing as the use of PCs grew. Better responsiveness was needed in systems development and in data access for improved customer service. Customization was easier through industry-specific point solutions. Costs seemed to be reduced through smaller implementations.
Now client/server technology has lost much of its luster. Who has enough staff and time to take care of all the servers? What's more, middleware costs now exceed the investment in the network itself. And infinite points of failure can make even the youngest technical consultant go gray.
These agonies could not have been avoided, but they could have been eased. In the planning process, the downside-or the "counter-upside"-must be examined.
For example, in the mainframe environment, we applied security; we did backups; we applied changes carefully. Using mainframe disciplines to manage large servers that support several applications will cure many of the headaches caused by integrating multitudes of single-application servers.
What trend are we in now? One of the scariest new technologies is software "objects." Market hype says objects will encourage data sharing, code reuse, smaller components and quicker time to market.
As your organization embarks on software objects or any other technology, ask what made your last technical implementation successful. Think about what in the new technology makes it inherently weak. Those weaknesses and the former strengths give strong clues about the "over-the-horizon" technology.
Our industry is quite good at recognizing cracks and fixing them. But fixing takes time. The improvements will be in the next release, the next product, the next revolution. We can forestall some of the problems by not letting infatuation blind our own good sense.
-- Judith M. Umbach is the executive director for the Year 2000 for the city of Calgary, Alberta, Canada. She can be reached at [email protected]