The Permanent Software Crisis

Abstract

Computer pioneers tell us that at the turn-on of the first computer, before the first program had even started to run, before the Great Machine came alive, while it sat there silently staring back at its anxious creators, the question of whether the First Program was correct came up, followed almost immediately by the question of whether anybody can write a correct program and be certain it is correct. A few minutes later, while these two questions were being explored experimentally by the designers and builders, with their hands and test probes deep in the machine’s electronic entrails, emotional doubts about the competency of overpaid programmers were vigorously articulated. For all practical purposes, the three questions that swirled around in the first few minutes of life of the first computer remain unanswered to this day. They are:

• Is this program correct?

• How can correct programs be written?

• Why does programming cost so much?