primitives

Instead, you need an environment that is rich in primitives that have been implemented in a communication-efficient way.

So trying to keep language small may lead to different decisions based on which primitives you want to keep.

Since Blake's round function is based on ChaCha20, this also reduces the number of different primitives you're relying on.

A good thing too when the NSA started monkeying around with crypto primitives and national security letters.

Yes, you can provide primitives to allow the programmers to do it but this does make things difficult and not-reliable.

Add another object as flow control, and you have all the primitives you need.

Compiler optimizations matter even more on ARM than on x86 due to shorter pipelines, smaller caches, super-expensive synchronization primitives, and so forth.

Yes - the best way to build complex things is out of simple primitives.

I'm all for investing in better primitives, but do these kinds of libraries really help vs implementing them from scratch with language primitives inside your (inevitably more complex) domain code?

ClojureScript could be decent: easy setup (now), extremely simple primitives, unassuming features.

For example, keeping the amount of primitives small and core language simple.

These provide good patterns and primitives for automating most ops-ish things.

Anyway, I already outlined the argument in another of my comments, where I talk about set theory and integrals - the "axiomatic" primitives are different bunch than "useful" primitives.

I would only implement the low level "hardware" primitives in C, then implement the high level API in JS like Chromium's Blink-in-JS initiative [2].

If you get a good set of low level primitives exposed that are freely composeable, then you can stop adding "features", and let third parties develop those features.

I think PPK is missing the point being "features" being added, and rationalizing the design of the old Web APIs by making them a proper layered stack of primitives.