Those JPL guys aren't your average programmers. I was out there working last time the SuperBowl was in Pasadena, and had four tickets on the 50-yard line, which I'd won as a distinction for a piece of work at the aerospace company I then served. I didn't want to go, and, oddly enough, couldn't find anyone on the team who wanted to see the football game despite the fact that I would happily have given them the tickets. Those guys love their work and it's important to them.

Now, that was a long time ago, but I hear the attitudes and practices there have improved since then. At JPL, there's too much at risk to do that "seat of the pants" stuff. Rigor and dedication are the things most important to a good design. If the first thing you do when you sit down to start work is plug in your iPod, and if you first make sure your cellphone's turned on, then you probably lack those.

Those guys are serious about their work. They measure twice, and cut once, and they know how to design a test procedure that will verify that their hardware, firmware, and software work as designed. Their stuff is multiply reviewed and critically examined all the time. Still, it's possible to make a mistake. They try to guard against mistakes falling through the cracks, too.

If you do the analysis before the design, and document the design requirements, then design your test procedures, before you start the design of hardware, firmware, and software, and ensure that there's absolutely nothing included because it's "kewl" but not absolutely required, your chances of designing a bomb-proof system are pretty good.

RE