Rummy Vs. the Arguing Experts

I'm watching a retired general on CBS this morning talkng about how our current Secretary of Defense should resign.

I'm not going to get into the politics of this. But I will talk about the larger issue of debating experts.

I have no criteria to make a judgement one way or the other on this issue. I'm not a general, never was a general, and my knowledge of the armed services is nothing near what, say, my uncle had with 30+ years in the Navy.

So there's some generals who think Rummy should go. There's also some generals who think Rummy should stay. Both groups are on TV and the newspapers having at the other group.

Aside from the vague feeling that I've started living in some banana republic, since when do we have generals, retired or not, trying to make public policy? As Pat Buchanan said, it looks an awful lot like a coup of the armed forces during time of war. It's not anywhere near that an extreme case, but the precendent is terrible. What's next? Should retired generals have their own political party? Should they make recommendations on what the DoD budget should be? Perhaps during each armed conflict, we could have generals on live TV commenting on the action. Oh yeah, we already have that.

In my opinion, even in the corporate world, taking the current political issues off the table, experts have a professional obligation to their clients to come to an agreement on policy. It is not fair to expect non-qualified managers to argue detailed policy with people who do this their entire lives.

I was reading an article in Dr. Dobbs this morning. In it, the author is reviewing some software project management books. One book, I could tell, rubbed him the wrong way. Why, I don't know. The example he gave of poor judgement on the part of the author was -- the author felt you could never completely gather all of ther requirements for a software system. The reviewer knew this to be a totally false statement.

E-gads! Here we go again! On one hand, I have the author of a prominent software management book saying you can never have all of the requirements. On the other hand I have a reviewer for a famous software publication saying that idea is popycock. Poor management. Of course you can gather all of the requirements.

If I were an non-technical manager, and if these two guys were in my office, I would have quite a problem on my hands. What is it? Can I know all the requirements or not? In the end, I'd have to side with the guy who says I should know all of the requirements -- after all, that makes the most intuitive sense.

But I'd be wrong.

As a former president taught us, a lot depends on your definition of terms. For our purposes, let's consider "requirements" to be the "formal request from the users for features in a system"

Now obviously you have to have a list of stuff that your program has to do, or you ain't got a plan. So score one for the Dr. Dobbs guy. But if you think that your users are going to stop having requests for your program to do stuff simply because you've started to program it, you're smoking rope. Sure, you can say something like "After June 1st we are locking down the requirements." You can even have a ceremony. Get somebody to sign a piece of paper. Whatever.

But you ain't stopping people from wanting new stuff. That's crazy.

In fact, most modern process methdologies (RUP and XP for instance), plan on the users continuing to give you requirements throughout the lifetime of the system. Mainly because that's the way it works in real life.

In this case, both guys had a point, and both were making totally logical arguments. If you want to be a real pro in any area, you had better learn to understand both sides of "religious" issues like this one. Then you need to understand that confusing your manager doesn't accomplish anything at all. Professionals have an obligation to put the interests of their clients above their personal opinions and feelings. Sometimes if you get a big ego this isn't possible. If so, it's time to find some other work. Like TV commentator.