4 ms·

Well e.g. JSR is different to anything I've seen from the Bun team, so yes. I didn't mean Deno and Bun's core offerings are different, but that there are diffe

Well e.g. JSR is different to anything I've seen from the Bun team, so yes. I didn't mean Deno and Bun's core offerings are different, but that there are differences in what they're each applying time and effort to. I'm a Node user. I haven't yet seen a compelling reason to switch that justifies potential compatibility/ecosystem issues.

WorldMaker 2y ago
Deno 2 was a major release entirely about shoring up Node compatibility. That's arguably a big reason for why Bun seems to have a higher velocity right now of the trio. Deno put in a ton of engineering into Node/npm compatibility, it shows, and right now I would recommend trying Deno as having suprisingly few compatibility/ecosystem issues remaining. (Deno has not stayed entirely still, other new features were added outside of that compatibility effort, but Bun doesn't have Node compatibility as much as a goal today and so gets the fast mover award.) I started using Deno in hobby projects because I like the out-of-the-box defaults a lot better than Node (deno.json is a lot simpler than a lot of the cruft that package.json has acquired, but also includes more things in one place like out of the box eslint support [deno lint] and prettier-equivalent [deno fmt]). Also, Deno Deploy has a generous free tier and that's a healthy incentive for hobby projects that want a modest database (Deno KV) and basic task processing queues.
- crabmusket 2y ago
  I also used Deno for some side projects before 1.0 and I really like it. Especially its focus on web stanrards, and yes, being an all-in-one tool. But I don't have the time or inclination for a lot of side projects these days. I actually feel much more warmly to Deno than Bun, and would use it if I felt like it made sense. I tried to avoid advocating either way in my former posts in this thread. But regardless of my personal feelings, at work pragmatism rules the day.

WorldMaker 2y ago

Deno 2 was a major release entirely about shoring up Node compatibility. That's arguably a big reason for why Bun seems to have a higher velocity right now of the trio. Deno put in a ton of engineering into Node/npm compatibility, it shows, and right now I would recommend trying Deno as having suprisingly few compatibility/ecosystem issues remaining. (Deno has not stayed entirely still, other new features were added outside of that compatibility effort, but Bun doesn't have Node compatibility as much as a goal today and so gets the fast mover award.) I started using Deno in hobby projects because I like the out-of-the-box defaults a lot better than Node (deno.json is a lot simpler than a lot of the cruft that package.json has acquired, but also includes more things in one place like out of the box eslint support [deno lint] and prettier-equivalent [deno fmt]). Also, Deno Deploy has a generous free tier and that's a healthy incentive for hobby projects that want a modest database (Deno KV) and basic task processing queues.

crabmusket 2y ago
I also used Deno for some side projects before 1.0 and I really like it. Especially its focus on web stanrards, and yes, being an all-in-one tool. But I don't have the time or inclination for a lot of side projects these days. I actually feel much more warmly to Deno than Bun, and would use it if I felt like it made sense. I tried to avoid advocating either way in my former posts in this thread. But regardless of my personal feelings, at work pragmatism rules the day.