Post by @MaddieM4@raphus.social

2026-11-07 03:44 UTC

One of the things really causing trouble for the Prone project (but great for my journey as a programmer, in C and in general) is that it's been helpful for refining my idea of what I want to make and how I want to make it. Which has led to me redoing work a few times in order to change the architecture. I really don't mind doing that. Redoing work isn't something I consider a failure mode, I actually think it's really important for the health and quality of a software project. If you aren't rewriting, you're probably just *emitting.* But it does superficially delay the project, even if every change in philosophy makes the total thing easier to write. That, I think, is a difference from traditional eternal project limbo. If you keep increasing the scope, you're probably not ever going to be done. If you keep decreasing scope, you might be going in the right direction.

Replies (1)

@MaddieM4@raphus.social 2026-11-07 03:57

To make this a little more concrete one step at a time in this thread, when I started out, I placed a really high value on performance. I was very offended by the wastefulness of modern code. And to a lesser extent, I cared about conciseness, but up to a point, conciseness and performance go hand in hand. But what about when the two are in tension? The further I've got into this project, the less I've cared about absolute performance, and more about permanence, conceptual simplicity, and robustness. I haven't stopped caring about performance, these are still values of mine, but watching the state of the world, and talking to other people online - seeing other people's good points and hearing their convictions - has shifted the proportional weight I put on these values. I am really troubled by things that make programming less accessible and less auditable. I'm more concern about wasted *scope*. My vision of ideal computing is inching closer to this: https://retro.social/@permacomputer/117572804972551071

Open ##5496037