lacasito25

u mean worktrees

in typescript we can do this

let res res = op1() res = op2(res.op1) res = op3(res.op2)

type inference works great, and it is very easy to debug and refactor. In my opinion even more than piping results.

Javascript has enough features.

Considering that you are bundling an entire runtime not meant to be installed independently on other computers, 67mb isn't that bad.

Go binaries weight 20mb for example.

I think they probably know that, this is alpha software, no need to be condescending.

why is having those complex flows done in a graphical UI better than just in code, where you can easily version them / test them?

I mean, you make S3 look easy compared to that

> Like materialism — the view that I sort of work with, the idea that consciousness is a natural phenomenon and is somehow a property of material things like brains and bodies. That itself is probably not testable

So, then, you "believe" in materialism, even thought you cannot prove it.

There's a more obvious theory than materialism: everything is consciousness. You create material things, they do not exist outside of you.

That looks more of a documentation problem, rather than an HTTP problem.

You just document that that POST endpoint doesn't actually modify data.

A great example of this are OpenAI completions.

here I am, using lit, dealing with OOP

There's the REMO framework, from daveshap, but I've yet to try that algorithm.

one can hope that with web assembly we can get closer to that dream, and eventually just remove the whole web browser.

How much money you think gpt3 training costed?

The project looks extremely messy, so many packages inside packages. It feels like java, but less performant

But what if "make testable hypotheses and then invalidate them experimentally" is wrong to begin with?