I'm optimistic that the next 12 months is going to a very special time for JS performance ⚡ (This applies to developer workflows & end-users.) Several large-scale independent multi-year language, runtime, tooling, & library investments are approaching the payback stage all around the same time. - ThreadSky

robpalmer.bsky.social • 27 days ago

I'm optimistic that the next 12 months is going to a very special time for JS performance ⚡

(This applies to developer workflows & end-users.)

Several large-scale independent multi-year language, runtime, tooling, & library investments are approaching the payback stage all around the same time.

Comments

robpalmer.bsky.social•27 days ago

At the JS language level Lazy Evaluation (`import defer`) is my personal favorite performance-related initiative because, just like `export defer`, it speeds up program execution time.

There's also require(ESM), Node's built-in TS, TS Go, Rolldown (plus the whole Oxc family), Turbopack, RsPack...

robpalmer.bsky.social•27 days ago

These foundational capabilities (and many I have not listed!) will act like rocket fuel for the @e18e.dev initiative that eagerly leverages the modern wave of tools, features, practices & processes to increase the efficiency of the whole JS ecosystem 👍

pi0.io•27 days ago

We will definitely try to utilize best of this opportunities across @unjs.io too 🚀

robpalmer.bsky.social•27 days ago

Of course! It's easy to see that UnJS shares the same values - and has done for a long time 😉

pi0.io•27 days ago

Since adopting import defer takes time as new language syntax, would you think would worth to advocate early support it in parsers? (specially oxc but also Node.js/SWC, kinda wish their typescript parser as OXC based too..)

robpalmer.bsky.social•27 days ago

`import defer` is business-as-usual for a language feature and shouldn't skip any regular processes that tools, runtimes & engines put in place.

There are tools & parsers that responsibly ship support for Stage 3 JS features today. So to the extent that's missing, I'd advocate for filling the gap.

robpalmer.bsky.social•27 days ago

The other side of the coin is that it would be risky to fully delay the tooling work for features until Stage 4 (and I'm not suggesting that ever happens) as the work itself can help uncover problems or improvement opportunities in the feature itself. So the feedback loop is valuable.

calebmer.com•27 days ago

I’m interested in making `import defer` the default for everything in our codebase. At previous companies I’ve seen 1s bundle execution times which I want to avoid!

But I don’t love namespace imports stylistically. How do you feel about a compiler plugin that turns all imports into defers?

martinszeltins.bsky.social•25 days ago

Rolldown will change the game!

chris-breuer.me•27 days ago

Yes, Stacks 1.0 is nearing, too, after years of development! 🥳

Comments

Posting Rules

Reply