Friday, June 24, 2005, 11:15 PM

Long-Lived Software

Been busy with work. Still am... but I did have some time to do some reading and thinking about the nature of systems we build.

To date, most of the software we build do not last very long. Heck, most movies have longer lifespans than most software. The people at the Long Now ( suggest that we should think about the long term impact of the systems and constructs that we create. The net is here to stay, and because the shorter-term thinking of the builders of parts of the net (e.g. IP address spaces), we are stuck with significant limitations.

Dan Bricklin has written about long-term engineering as it applies to software (, Bricklin emphasizes fact that the structure and culture of a typical prepackaged software company is not attuned to the needs of societal infrastructure software. He proposes that software engineers should learn from civil engineering.

As we embark on designing what could be the future of a permanent fixture in cyberspace -- the identity infrastructure -- we should be cognizant of the long-term impact of our actions and designs.