IMO, deno's approach was bad as it was reinventing the wheel, so one had to relearn. And then they brought package.json which they said they wouldn't. This again got people to unlearn and relearn things.
Bun, on the other hand, acts like what Typescript is to Javascript. It's just feels like superset of Node, instead of completely different tool.
It's almost 100% compatible with node but faster. A lot faster. So no need to learn anything but few cli commands. For example bun run dev instead of npm run dev.
I'm trying to get my work to switch to bun but we have packages in a private AWS codeartifact repo. Does it support this? I tried to use it with our npmrc file but it couldn't install those packages.
They deleted their tweet (or I can’t find it because I refuse to sign into Twitter on my phone) but when oven first started hiring they were extremely condescending towards the idea of work life balance. It’s a startup perpetuating the idea that startup employees need to cut themselves off from their lives and focus on work for minimal benefits - you know, until it “takes off”.
It caused quite a stir when it was posted, I’m surprised it’s so forgotten now.
Deno is still around and is even actively used, you have to use it if you want to write a Supabase edge function, for example. But it's not used in mainstream development from what I can tell, it just never took off because it's a very large idea shift from Node that requires a decent sized learning curve to figure out. The benefits are also not enough that it's worth re-learning how to write server-side JavaScript. If you wanna write server-side JavaScript, Node is good enough that it's not worth re-learning.
Still though, Deno is fairly obscure from a mainstream development perspective, and that's what I wish on Bun.