Teach your toddler binary in just 10 steps

Correction: we did not get quite that far. You can import a JavaScript module source without executing it, transport it, and execute it elsewhere, but not yet parse and compile it from text. In any case, another cog is in place.

RT @MetaMask: We're literally building a moat of lava to keep MetaMask users (and the entire web) safe. It's called LavaMoat, and it helps…

RT @naugtur: A Phish on a Fork, no Chips. One more thing to beware in the world of software supply chain risks. Read if you care abo…

OCapN’s data model is slightly more expressive and less ambiguous than JSON’s. Numbers are either float or bigint. Strings are strictly Unicode. Supports byte arrays. And, critically, it can embed references to remote/local promises and callable targets.

RT @Raynos: I’m going independent for next year. Just kicked off searching for medium term contracting engagements. If anyone has backend /…

@creationix If I were doing this in JavaScript today, I would model an HTTP Response something like: type Response = Promise<{code: number, message: string, headers: Record<string, string>, content: AsyncIterator<Uint8Array>, trailers: Promise<Record<string, string>>}>

@creationix The local idiom for readable stream should ideally be founded at the language like async iterator of array buffer or typed array, not coupled to an I/O idiom, or at least lifted from language features. Middleware should only ever see the fully lifted type.

@tlrobinson Sir, this is a Wikipedia.

Any method that reports to stdout is unsuitable for use by libraries to report diagnostics. We can rely only on warn and error to be web/node portable. Groups are useful to bundle all of a library’s diagnostics and reduce web console noise, but can’t be used on Node.js.

As a fellow tech-conference-absurd-hat-wearer, I admire @JeetRaut’s commitment to the bit.

RT @robpalmer2: ECMAScript excitement 😉 Congrats to @marksammiller on advancing Immutable ArrayBuffers to Stage 1 at @TC39 today 🎉 It pro…

@SeaRyanC @jarredsumner I have not looked closely enough to say one way or the other or assess all the avenues available to the shared structs proposal, but one will have to be found that does not erode the possibility of mounting a defense for code that interacts with shared structs at the edges.

@SeaRyanC @jarredsumner And when other programs interleave within yours, the author also has a measure of control over how those programs can interfere. So, does the feature in question break or not break past, present, or future plans? Does an unsafe block give them a place to mount a defense?

Please take a look at for more about sandboxing your sus node_modules and plugins. Neither modules nor promises are finished. There’s a lot of potential for both.