In between my other reading, I'm still looking at Software Creativity 2.0 (I also wrote about it here). The third chapter, "Optimizing vs Satisficing", is all about knowing how far to chase perfection. Robert Glass comes down squarely in the camp of 'good enough' software. Then I hit the third essay in that chapter "Ad Hocracy" — 'What's this doing here?' I asked myself. It doesn't really fit cleanly with the first two sections. It's a rant of a different color (or something like that).
I read through the essay carefully. I kept hitting up against the difference between this semantic battle to restore the good name of ad hoc approaches, and the more noble cause of shipping 'good enough' software. Then it hit me, right at the end. He's really talking about the same thing. It's the same fight, just on an even larger scale. The hunt for the perfect, generalized answer is the ultimate in optimizing — when we can give that up and write software that 'satisfices' by solving the problem at hand, we can ship 'good enough', ad hoc software.
It's essays like this that help me understand why the first edition of Software Creativity became a cult classic. I'm really looking forward to getting my hands on a 'real' copy of Software Creativity 2.0.