Why make software? | Sean Voisen

If you want to leave a lasting legacy—some durable remnant of your brief existence that will ripple forth into the future and leave an imprint on hearts and minds of future generations—you might be wise to dedicate your life to pursuits other than the creation of software. Writing, painting, architecture, scientific research, building a business, perhaps? If you’re talented, dedicated and lucky, all are more likely to get your name in the history books and solidify a legacy more than making software.

Software is ephemeral. It is building sand castles in the air, forever shifting, changing, eroding. The life span of most software is at best a few years to a decade, and much of what endures is constantly evolving, built and maintained by teams of humans who remain forever hidden to the end-user, a rotating cast of otherwise invisible programmers who come and go.

This battle with entropy has nothing to do with the nature of software itself, and everything to do with its supporting socio-technical systems. With the right storage medium, code can persist in stasis nearly indefinitely. “Bit rot” is thus a bit of a misnomer, because while it is true that electronic signals are ephemeral and decay, code doesn’t rot. Instead, the systems that support it—both hardware and software—change out from underneath it. Dependencies change or become deprecated, remote APIs start responding with 404s, operating systems update, CPU architectures change or are no longer manufactured.

Leave a Comment