I'm OK with not knowing Deno's future

With Cloudflare's acquisition of Deno, people are expressing sympathy for those who, like me, have gone all-in on the Deno runtime. It will see 12 more months of security/bug-fix releases, and after that nobody knows what happens. Thank you for your concern; I'm going to be just fine.

Why I don't love Node.js

A lot of people are talking about Deno as if using it is chasing exciting bleeding-edge technology. They say Node.js is boring, but you can count on it. I don't have a problem with Node.js being boring, but being both boring and tedious is a problem. Much like a coding agent, I have limited context that fills up. Suppose I want to play with an idea of something to do in TypeScript. Node.js can run the .ts file, but for a nice editing experience I also need a language server and formatter. Node.js forces me to either temporarily pollute an existing project with my new toy code, or create a new npm project just to play.

Why I love Deno

With Deno, the tedium is gone. I can put a new TypeScript file wherever I want, and Deno takes care of everything. Its built-in language server and formatter mean I can edit the program with all the niceties I'm used to in a fully set-up project. The fact that TypeScript has to be transpiled for a JavaScript engine to run it is effectively hidden from me. I just program.

These days I'm often writing ancillary scripts for agent skills. The self-contained nature of Deno means that they can be in TypeScript just as easily as Python, or actually more easily because Deno handles package management more smoothly.

Deno follows boring standards

What people don't understand is that Deno is boring technology. That's why Ryan Dahl doesn't want to keep working on it. He wants to work on more exciting and original things. Deno is very standards-based. It already achieved compatibility with standard JavaScript and TypeScript; this is the boring part that I most like. Learning Deno is largely learning browser APIs, so my time is never wasted. This is great for me, but why should I expect Dahl to be excited about continuing to adhere to standards? That's unoriginal. Deno already has an unequaled TypeScript developer experience; there probably isn't any exciting news in its future on that score. Deno's roadmap, in Dahl's words, "has been sucked into the gravity well of node compatibility."

I'm going to be just fine

So here's the worst case for me: In a year, nobody picks up Deno development (unlikely). Then some new bug or security issue ends up affecting my code, and nobody steps up to fix it. I have to port my Deno stuff to Node.js or Bun. With a coding agent, that will take less than a day. Between now and then I intend to keep using Deno. The time I save due to Deno's convenience is worth a year of not knowing the future. And Node.js and Bun might catch up with Deno in some areas in that time. Migrating should only get easier the longer I wait.

Best case, Deno keeps going, and I keep enjoying it.