Assumed Audience: People interested in my writing app project, in software development in general, and in “winning slowly” (as it were).
Epistemic Status: Talking myself through this, to be perfectly honest.
I spent a little time this weekend working on getting the URLs for the web app version of rewrite working. I’m about 80% of the way done with it. For the last bit, I’ll be asking a bunch of questions of folks more experienced than me with Elm: I can see ways to solve a particular challenge, but I don’t like any of them. This is just part of the normal process of learning a new technology: things go a bit more slowly because you don’t already know what you’re doing.
I had the thought today — not for the first time — that perhaps I should drop Elm and switch to just doing Ember and TypeScript: technologies with which I’m deeply familiar and very competent. I could undoubtedly get the web app up and working faster that way, and I don’t have any particular interest in needlessly reinventing the wheel. As I mulled on the question on my run earlier, though, I came back to the same decision I have every time I have thought on this.
I’m building this thing in Elm on the web and with SwiftUI on iOS and macOS, even though it is definitely slower to get to market than just shipping an Ember-TypeScript-Electron app, because speed-to-market is not really a primary concern for me. I have the luxury of working on this on the side, and I want the result of my efforts in building it to be a code base that will last a decade or two. More than that: a code base I will be glad to work in for a decade or two.
When your timescale is not 6 months of runway but as long as it takes to make it good there are downsides: it can lead to paralysis, or the endless pursuit of unattainable perfection. But it also has the very great upside of being able to build something well. We don’t build highway bridges overnight, because they have to endure. Too often, we try to build software overnight, without considering how it, too, may have to endure.