Some weeks ago I started learning demo at my work. Their built-in support for Typescript and being developed in Rust seemed to be its strong points.<p>Now, there is no actual codebases I can look into, SQLite driver is broken, there are several ORMs that looked interesting but all of them are deprecated (I chose sequelize with MySQL), I’m also having disconnection errors when trying to do integration tests…<p>I feel discouraged and wonder if I’m a bad engineer (maybe that’s the problem) or deno is still in experimental phase.<p>Have any of you regretted choosing a cool technology? Did you used a different one or persevered with your decision? How did you keep working with that technology?
Node JS and by extension Deno is built on a house of cards.<p>Deno is simply not ready for production. You can import NPM modules, but they aren't guaranteed to work.<p><a href="https://docs.deno.com/runtime/manual/node/" rel="nofollow">https://docs.deno.com/runtime/manual/node/</a>
Deno added Node compatibility later on. So while Deno itself (JS which is V8 anyway and the native Deno APIs like http) is probably stable (it is used as serverless runtime, e.g. from Supabase, of course Deno itself and Netlify IIRC), the Node compatibility is the problem here. Yeah, this is frustrating. Now we have Bun which tries to be a drop in replacement for Node. It you need NPM modules, you should probably just use Node. Or some other language.
I'm not, but am looking forward to everyone getting disenchanted with Go so I stop seeing yet another startup pick it for something it's bad at and then struggle to eke out performance and scale the project.