Premature Optimization

In programming, there has long been a warning in computer science to avoid premature optimization. Donald Knuth called it “the root of all evil”. I find myself thinking about this all the time—not so much while programming, but when I’m thinking of spending money on myself, or telling people how to spend their money on me, as for birthday or Father’s Day gifts.

I’m at an age now where I have everything I would ever want. But…everything I have could still be a little bit better. To wit:

  • I have awesome headphones that I love. I want better ones. And different ones.
  • I have a home server that is underpowered, but quiet and extremely reliable. I would love one with enough power to could run virtual machines, but I don’t really need it.
  • I have a clicky mechanical keyboard that I love. I want a better one—that lights up, unnecessarily, or has colorful keycaps.
  • I have an Apple wireless keyboard for my Mac. I want to replace it with the Apple Magic keyboard, even though I already have an Apple Magic keyboard for my iPad.
  • I have a Series 1 Apple Watch that I love. I would love, even more, a Faster Series 3.
  • Let’s not even talk about iPads and Macs.

These few things are some of my material obsessions. What they have in common, for me, is that they have all been satisfied by things I already own, upgrading to newer or better versions would cost a lot of money (way more than anyone would spend in a gift for me), and the upgrade would be only marginally better than what I have, so I’m not sure if it would even make me happy.

Despite knowing all this, I can’t stop thinking about upgrading what I have to something better. I always want to optimize my experience with the things I enjoy. But, until the things I have break down and are no longer useful, it is too early to upgrade them. Doing so would be indulging in premature optimization, which be wasteful, which is “the root of all evil” to me.

Someday, my headphones will break, my keyboards won’t be compatible with my computers, my server won’t support the OS I want to run, and Apple won’t support my hardware anymore. That’s when I will upgrade these things—after I have extracted every bit of their value. Until then, I will just daydream about, and feel a little guilty about obsessing about, premature optimization.