Ted MacNEIL wrote:
I don't know how successful it was, but years ago we used function points. Rather than counting lines of code, we would count new functionality (or modified function).
Last millennium I was a contractor at a government agency that used function points to evaluate programs and coders. In the most extreme case, I had a program that performed twelve distinct, but logically related functions, based on a PARM value. It was scored accordingly, whereas I did perhaps 10-20 percent more work than on a simple program; only the testing was more elaborate.
In forty-odd years, I have yet to see a meaningful, reliable metric. From the business aspect, the only thing that matters is that when something is needed, it can be finished in time. A corollary of this is that management should make provision for goofing off; programmers who have time to experiment tend to be more productive. I first ran into this when my service bureau employer acquired 2260s for a customer, and we got to play around with them. By the time we were comfortable with them, the company used our experience to win a government contract to convert a batch package to interactive 2260 use. Perhaps my favorite example is the computer operator who wrote a football program, and eventually wound up as an IBM manager.
Gerhard Postpischil Bradford, VT ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO Search the archives at http://bama.ua.edu/archives/ibm-main.html