terra VS llvm-project

Chapel has at least several full-time developers at Cray/HPE and (I think) the US national labs, and has had some for almost two decades. That's much more than $100k.

Chapel is also just one of many other projects broadly interested in developing new programming languages for "high performance" programming. Out of that large field, Chapel is not especially related to the specific ideas or design goals of Mojo. Much more related are things like Codon (https://exaloop.io), and the metaprogramming models in Terra (https://terralang.org), Nim (https://nim-lang.org), and Zig (https://ziglang.org).

But Chapel is great! It has a lot of good ideas, especially for distributed-memory programming, which is its historical focus. It is more related to Legion (https://legion.stanford.edu, https://regent-lang.org), parallel & distributed Fortran, ZPL, etc.

Terra is the language you're looking for: https://terralang.org/

Wow, amazing stuff. I love Lua, it was how I learned programming as a kid. Coincidently from the same world as the author. Open Tibia.

The author made a custom client (https://github.com/edubart/otclient) for the game that is still very much in active use by thousands of players. He's a very skilled developer.

Great to see AOT typed Lua, I know of the other solutions: Luau, Teal, TypeScriptToLua, Terra, etc., but this one is my favorite so far.

Love the simple compilation to C (and WASM support via Emscripten). Though Terra's JIT is enticing and good replacement for LuaJIT, this is for embedded systems, it's a good replacement for Lua PUC-Rio.

The World:

- https://luau-lang.org/

- https://terralang.org/

- https://github.com/teal-language/tl

- https://typescripttolua.github.io/

Terra is a language that can also do that, and uses Lua as the metaprogramming language. Types are just Lua values.

But unfortunately, there's a lot of work left kind of half-baked so using the language is a pain... if someone invested a lot of time to make Terra work properly and added some tooling around it, wrote proper docs and so on, it would be a really interesting language.

https://terralang.org/

(I am slightly familiar with a language called Terra (https://terralang.org), which couples C with Lua, where the Lua is basically used as the metaprogramming layer ... sort of like comptime in Zig. And making an SOA data structure is the kind of thing you could do in Lua in Terra. So that was partly the basis for my question).

> why can't the hardware designer do something simple and clean

If it was easy, it would not need firmware in the first place. Firmware is there because people expect certain features and quality of life. See softmodems.

> write some assembly (without abusing the assembler preprocessor...)

You want https://terralang.org/ and not "just C"/"just Assembler" instead ?

I forgot to say that I use this llvm build. Just download the release and point the clangd server to it.

Xtensa support (esp32). Will be interesting how this will be for Rust and Zig support for esp32

https://github.com/llvm/llvm-project/tree/main/llvm/lib/Targ...

https://www.phoronix.com/news/LLVM-Xtensa-Backend

https://github.com/espressif/llvm-project/issues/4#issuecomm...

https://github.com/ziglang/zig/issues/5467#issuecomment-1465...

Hopefully, in the future the installation will be simpler, as we are trying to upstream our LLVM changes (first 10 patches are already accepted!), and once we manage to upstream LLVM changes we will proceed with upstreaming our Rust fork changes.

The main discussion can be led back to this GitHub issue and this comment; fortunately, we don't need to build espressif's llvm fork anymore as they supply the clangd (this is the language server we need) and you can find the zip here. I'll briefly list down the steps required to set up vim with clangd to take advantage of clangd's features (auto-completion, linting, code refactoring ...)

In my case, it's pretty annoying that the Xtensa platform doesn't have official LLVM support. It's in progress but going very slowly.

The link to the espressive issue trackers: https://github.com/espressif/llvm-project/issues/4

Xtensa. They've got a fork of LLVM that supports it that they're working toward getting upstreamed. The community has a fork of rustc that uses it (and a quickstart crate) while we wait for it to get upstreamed.

https://github.com/espressif/llvm-project/issues/4 is a good example why updating llvm isn't easy and takes a lot of time.

Just to clarify - Gambit, Chicken, and Carp all compile to portable C.

I hadn't realized LLVM mainline doesn't support Xtensa. I'm surprised.

D does support Xtensa via LDC (https://forum.dlang.org/thread/[email protected]...). It looks like GDC also nearly supports it, requiring only a minor patch at present.

A functioning LLVM backend does exist (https://github.com/espressif/llvm-project/issues/4) and might be making very slow progress towards being merged. A quick search shows that it works for Rust. I suspect (but don't know) that it might work for Terra as well.

There's also the LLVM C backend (https://github.com/JuliaComputingOSS/llvm-cbe) but I've no idea how efficient such an approach is when applied to real world embedded tasks.