sorry, not yet 🙇‍♂️ we had a small thread here https://discord.com/channels/817392104711651328/1103003569654865941/1103003569654865941 and i created a tracking issue here https://github.com/garden-io/garden/issues/4158 it has already got quite a few reactions - people do want it 😅

Turns out Node 20 is not at all supported by pkg, and I hesitate to guess when that will happen. We might consider this instead then: https://github.com/leafac/caxa Some pros and cons, but worth an experiment imo. I'd be curious to see performance impact, startup time etc. The latter could be affected substantially by switching to Yarn PnP (which does not jive with pkg), or outright bundling the whole thing, because it drastically reduces I/O on startup. Thinking ahead, the general direction of making a Go CLI frontend for

garden serve

operating as a daemon (exempting interactive commands like

garden dev

, where startup time is less of a concern to begin with), will mean we care less about startup time, but we do have to think about it in the meantime. But if migrating to caxa results in a proportionally small startup time delta (ignoring the very first startup which we could address at install time), it may well be worth it given the substantially simpler architecture. By all means hold until post 0.13 launch, but I'd really like to probe this after.

made a small attempt today at node 20 to see what kind of effort would be required. on the plus side, looks like there aren't that many breaking changes in 19 and 20 on the minus side, as was mentioned above, we're blocked on this until

pkg

and

pkg-fetch

add

node20

support for the

build-dist

step various other tests seemed to pass though! https://github.com/garden-io/garden/pull/4346

Yeah, I suspect we'll need to later investigate alternatives to pkg, but that's a non-trivial project.

added a tracking issue https://github.com/garden-io/garden/issues/4428