Sunday, October 31, 2004

Systems that are "good enough"

Bruce Sterling wrote a note about Clay Shirky's article about evolvable systems and the Web. The article is old, but definitely relevant to what I'm writing here.

The systems we all use daily, whether they be Internet, Web, television, telephone, or electricity; the ones we just assume are there (until they're not) - these systems all evolved over time, until they were "good enough" to be widely adopted, but not perfect. The key is that they continue to evolve - when something no longer is good enough to do what we want, extensions are created, grafted on (sometimes with bubblegum or duct tape), and we move on. Sometimes, we come back and make the extension work better, sometimes it evolves until it seeems part of the original fabric of things, sometimes extensions die out when something better comes along.

If you're old enough to have had a personal computer in the 80s (or even back in the 70s), think of your first personal computer, then look at everything on your current one. It's still a computer, still basically the same system. Possibly, it's still running an Intel architecture. But look again, or think again, and you realize how much it has evolved over time. Try to remember all of the extensions as they occurred, and the proposed extensions that didn't make it - it's an eye-opener in many ways.

No matter how the changes occur it's an evolutionary, rather than revolutionary, process that produces results that stand the test of time. The changes at any given moment are just "good enough" to do what we're trying to do, but with plenty of room for improvement. When you see those changes occurring, you're in a good place. When someone tries to give you a "quantum leap" or "completely revolutionary" change - watch out, that particular technology is possibly going to be a backwater eddy to the main stream.

No comments: