Why JavaScript?

Today in the inquiry into the UK Post Office IT scandal that led to >700 baseless prosecutions...

🚩 The Post Office 'Head of Legal' stated they are 'not sure' if the standard of proof in a criminal trial is that a jury had to be sure of guilt.

postofficescandal.uk/post/what-hugh…

The nice part about this is you can write nice functional code that iterates on collections (map, filter etc.) with hand-optimized-level performance.

Every time I use map/filter in JS instead of plain for loops I feel like voluntarily giving up performance.

Very interesting. This could be a good way to speed up CPU intensive tasks in some JS libs without adding a native dependency. It may move the line to switch away from pure JS a bit further.

Beyond Wasm, MoonBit now supports #JavaScript backend!🤩
Up to 25x faster than native JS, and nearly 8x faster than Json5🚀🚀
How it’s possible? Read our latest blog👇 moonbitlang.com/blog/js-support

Slack chose Deno to build its next generation platform because:
✅ native TypeScript & web standard APIs support
✅️secure-by-default,👍for enterprise apps
✅️can be compiled into a portable, self-executable binary

deno.com/blog/slack-sav…

Build tools embracing ESM as their own method of deployment is a key step forwards for broader ESM adoption.

It promotes 'dog-fooding' or 'pipe-cleaning', meaning learning lessons that will feedback into a better migration experience for libraries & apps using those tools.

Incredible. This wouldn't be possible without the previous work Jake et al have been doing in the TypeScript codebase. But Joyee's unlocking require(ESM) in CJS will greatly speed up the ongoing ESMification process. We'll be able to publish Vite as ESM only in a few majors.

After many import rewrites and a couple of tweaks, it looks like it may be possible for TypeScript to go ESM, including the public API, without breaking CJS once --experimental-require-module is unflagged (assuming github.com/nodejs/node/is… makes progress). Very very very cool.

This brings up some great questions about a confusing topic, but the advice Brandon found in his research is bad advice! My full explanation via blog post: blog.andrewbran.ch/default-export…

This is the primary action advised by Andrew Branch.

It helps folk who are compiling their modules to CJS. And it helps folk who ship ESM.

So it's just good all round.

↗️ Guidance on default exports in CommonJS libraries by Andrew Branch

'0.1 + 0.2 != 0.3? JavaScript doesn't know simple maths works!' It's obviously not JS's fault, it's how floats work and most languages do the same.

However, did you know that my colleague Jesse is working on a 'decimals' TC39 proposal to fix it?

Wes Bos We're kinda blocked on not being able to support it in JS, so the roadmap is available here: github.com/tc39/proposal-…

The JS runtimes we have internally use crash telemetry in dev & production. So infrastructure developers get alerted quickly to crash trends that can be sliced & diced by versions & environments.

A real-life data-driven approach to quality is vital IMO.

I'm always excited by a TypeScript release. This one is epic, and what's all the notable is how much community involvement there was in it. I loved seeing the shout outs to Titian Cernicova-Dragomir 💙💛, Dan Vanderkam and others in the release post. Recognition is so important and so appreciated ❤️

bun.report is bun's new crash reporter

bun.sh/blog/bun-repor…

Do most people really like `string[]` more than `Array<string>` in TypeScript ? Why is that? I have some quite convincing arguments for the Array syntax: ⬇️🧵