People often respond to my FIST (Fast, Inexpensive, Simple, Tiny) presentations by saying that “Everyone agrees with this," by which they seem to imply this stuff is almost too obvious to say.
The thing is, that’s not exactly true.
Sure, everyone agrees in principle to the *objectives* of FIST, such as reducing the cost and time involved with developing new systems. But when it comes down to actual *implementation,* we quickly encounter friction. And it’s the implementation that’s at the core of FIST.
Let’s take a look at a few specifics, shall we? In my FIST presentations & articles, I say simplicity is desirable, a sign of sophistication. Adding more features, functions & widgets makes the system worse, not better. But take a look at almost any piece of consumer electronics or any weapon system, and you’ll find it’s chock-full of shiny bits and features, apparently added in the belief that more is better. In fact, when you actually analyze the system, you often find it’s full of extra features, which:
a) are unused / seldom used,
b) increase the cost of the system,
c) increase the complexity of the system,
d) reduce the reliability of the system (by creating more potential failure modes, for example),
e) make the system harder to learn, use & maintain
f) all of the above
I could go on, but you get the picture.
Similarly, FIST argues that constraints foster creativity, and thus small budgets are good. But as soon as a program runs into a technical problem, the same people who supposedly agree with FIST end up asking for more time and money. Or consider the huge pressure on military program managers to “keep your obligations’s & expenditures up,” (i.e. spend all your money) for fear of not getting as much money next year. Clearly, the environment does not reward thrift. Any DoD program manager who turns in money at the end of the year is likely in for some harsh words from the finance folks. Spend, spend, spend! But yes, we agree that FIST is a good idea… as long as we don’t make things simple and as long as we spend a lot of money.