Developer's Cut

Jan 11 2025

I was watching HBomberguy’s video on Director’s cuts a few days ago, when my mind discovered a connection to something else I’d read, as it tends to do when I’m marinating on the couch after dinner. Hbomberguy references an article ‘Uncertain Glory’ by Greg Solman, published in Film Comment magazine, 1993. The article is a terrific read; here’s my favorite passage (and most relevant to my larger point):

Viewing film art as perpetually subject to update and correction, political or other, says something about the director’s relationship to his work…Director’s cuts are both futile and wrong-headed. In a way, they redress artistic grievances by treating movies as mere product. In this country, an inalienable right to change as an artist doesn’t extend to changing art and, therefore, art history.

I don’t think the argument is that Director’s cuts should be outlawed as a concept — rather, treating cinema as subject to revision and change (which can be marketed) has consequences for the creative pursuit of film-making and how the audience responds to the art-form. Which brings me to this reply by John Gruber to Ken Kocienda’s post on Mastodon:

For as long as I’ve known him — and I’ve known him going on, jeez, almost 25 years now — @brentsimmons has said his favorite part of programming is deleting code. I think that’s perhaps the fundamental thing about programming that makes it unique as an art form. No other creative medium I can think of has that quality.

While this reply is about deleting code, Kocienda’s original post describes the way most (good) software developers view their code. We learn not to grow attached to our code, and to focus on constant updates and refinement. But what does this say about our relationship to our work, as Solman raised about Director’s cuts?

I’m not going to argue that programming isn’t art; I’ve already written about how the Is this art? argument isn’t very interesting. However, two things are true at once:

  1. Programming is fundamentally revisionist — you have to maintain and update your code-base or it usually won’t work sooner or later. There’s always bugs and inefficiencies to fix. A software project not being actively maintained means death. Programming might have artistic qualities and it is definitely a creative pursuit — but the intended product of your program has to work and be stable for some time to be useful.
  2. However, the fact that programming is fundamentally a revisionist creative pursuit means there are consequences to embracing it whole-heartedly. Big Tech releases half-baked software all the time under the Beta label. Video-game companies release unfinished video games and market it as Early access. Just because software is amenable to updates and refinement doesn’t mean our standards for good software releases should go down, but there’s where we are now.

Drawing a connection between the unintended consequences of Director’s cuts and revising code gives us a new lens to understand how we approach building software, but it shouldn’t be taken as a directive to update or revise code less. I think knowledge of the unintended consequences of the nature of one’s creative pursuit or medium is valuable in of itself. It helps us make better decisions when writing code and building software.