Computer Generalist
5833 Ross Branch Road
P.O. Box 262
Forestville, CA 95436-0262

Date: Tue, 22 Feb 2000 19:41:35 -0800 (PST)

Mr. Rod Welch
The Welch Company
440 Davis Court #1602
San Francisco, CA 94111 2496

Subject:   Advanced PM Program at Stanford

Aha! You say, "they feel that "communication" is their strongest talent."

But I know that if you ask people quite seriously to rate their own ability to communicate, virtually every one of us will place ourselves in the "above average" category.

Virtually every week I find myself explaining to another friend that thirty percent is a good failure rate for simple "Honey, would you bring home a loaf of bread" type messages. Despite the fact that every one of us believes that those get through ninety nine percent of the time, the true value is usually between thirty and eighty percent. Even ninety percent is going to be very rare, and then only for folks who will be thought of as fanatical.

Imposing a systematic confrontation with reality is going to feel bad for a lot of folks who would have to learn to live with a less rosy picture of themselves.

If you want evidence of another kind, consider the checkpoints in a multi-month project. The guys doing the work are going to try as much as they can to reduce the number and magnitude of checkpoints. They'll say, just tell me what you want, leave me alone for 3 or 6 or 18 ticks and come back when it's done. You're gonna love it.

You won't.

This is how you get giant failures. It's like setting the boat pointed precisely at New Orleans and tying the rudder. It is necessary to make many, many course corrections and Gilb, Tom, in Principles of Software Engineering Management explains in considerable detail the wealth of dividends you get from resisting the urge to skimp on checkpoints. He urges you to deliver a complete working system, better than the old one, every two to five percent of the project resources, including calendar time. One major result is a complete lack of major failures.

It's a big deal. Hardly anybody gets it. If something lets folks get what the truth is and keep on truckin' that will change the future in major ways. Your SDS or Gilb's Evolutionary Delivery both focus the mind on fixing problems before they escalate. But both require that you accept reality, personally. Very different from taking reality personally. More like taking on reality, personally.

You hear?



Dick Karpinski