Unlike the '80's, nowadays, there are other things that are as important as actually knowing how, and having the needed discipline, to generate a sensible program. Today's issues are not new, but they're getting less attention than they once got. These include simple questions such as, "Which tasks can reasonably be sujected to full or partial automation"? Too often, I see folks attempting to "computerize" tasks that are insufficiently rigorous to permit them to be automated. Programs become needlessly complex when the computer is required to do the analysis that the system designer should have done, and that leads to "cut-and-try" approaches that most often fail over the long run. The real problem doesn't seem to me to lie in the dearth of programmers or engineers, though there's little overlap between those two, but rather in the dearth of competent analysts. Kids in high school can learn enough to program a well-defined task. However, they're not trained in high school to analyze a problem before attempting to solve it, and the colleges don't really do that either. Consequently, if a person hasn't learned to "be an engineer" before he's asked to do a job, he probably can't do it. Engineering is a discipline that's largely based on a personality type. That type is bent on solving problems with what's available to him/her. Not everybod comes from there. Dick ----- Original Message ----- From: "Douglas Quebbeman" <dhquebbeman(a)theestopinalgroup.com> To: <classiccmp(a)classiccmp.org> Sent: Monday, August 20, 2001 6:54 AM Subject: RE: SemiOT: Mourning for Classic Computing