Niles Salter

Do you think instructions (with commands) for how to do this could be added to the README?

You might want to take a look at [my fork of the F3X Try library.](https://github.com/Validark/RBX-Try-Library/blob/patch-4/TryLibrary.lua) The [documentation is available here](https://rostrap.github.io/Libraries/Try/). I don't really like it though because it doesn't really...

Thank you for this contribution! I think it might be better if we don't even do Promise.delay() for an input of 0. It just seems unnecessary to insert it into...

The reason why it's all in one file for now is just because it's more convenient for me. I know I can objdump on my own machine but it's nice...

1. Yes, this implementation should be quite favorable towards embedded. - Reducing the memory footprint of the AST is one of my goals, which is another step towards supporting 100MB...

> As for my construction of the AST, I am not sure which technique I will ultimately prefer, but it won't matter either way to consumers of the AST. Actually...

> using Zig compiler on embedded (100-300MB RAM) is a potential use cases Could you give an example of representative embedded hardware? Do they typically have 64-bit trailing zero count?...

> I dont think (permature) optimization is worth the hassle, unless to prevent significant performance dropdowns. It is not premature optimization to think about the performance of your software on...

I just got done putting in a lot of hours learning SWAR techniques and implementing pretty efficient SWAR fallbacks. I have not committed it yet, as the code still needs...

Update: I tested out the tokenizer on a RISC-V Sifive u74 and it's ~1.5x faster than the legacy version! With some more work, I might be able to improve it...