axiom VS Vrmac

I'm creating a hobby operating system in Nim (early stages). After an initial attempt[0] I decided to start over[1] and document my journey in building it[2]. The focus for me is on learning low-level systems programming and enjoying the journey, rather than building something production ready.

[0] https://github.com/khaledh/axiom

[1] https://github.com/khaledh/fusion

[2] https://0xc0ffee.netlify.app/osdev/01-intro.html

> kernel developers do not allow third party runtimes in the kernel. Even meager Rust's "panic" runtime is a contentious

Much in Linux is contentious :-) which is why the module system is nice. A kernel module for C code requires no permission from Linux-core unless you need it distributed with the kernel (which, yes, might be required for "credibility" - but critically also might not). It may require many decls to access various kernel APIs, but those can be (semi-)automated or just done as-needed. So, Linux kernel policy is not so relevant (at best) which is what I meant by "no special support" (admittedly brief). Kernel coding is always a bit trickier, and you may need to build up some support code to make integration nice, though as well as decl generators.

> Can one disable runtime in Nim completely -- no GC, no exceptions?

To answer your question, and as discussed elsewhere in this subthread, Nim has many options for memory management.. only stdlib seq/string really needs automatic methods. One can disable the runtime completely via os:standalone and statically check that no exceptions are raised with Nim's effect system (and there are also both setjmp & goto based exception impls which may/may not be workable in Linux/BSD kernel module settings). As "proof more by example", a few people have written OS kernels in Nim recently[1,2] and there was another toy kernel long ago[3].

People have also written OS kernels in Go which "has a GC and runtime".[4] So, I acknowledge it's not quite the same example, but I also see no fundamental blockers for kernel modules.

[1] https://github.com/khaledh/axiom

[2] https://prosepoetrycode.potterpcs.net/2023/01/a-barebones-ke...

[3] https://github.com/dom96/nimkernel

[4] https://github.com/mit-pdos/biscuit/

I gave Rust a few chances, and always came out hating its complexity. I needed a systems programming language to develop a hobby OS[1], and Nim hit the sweet spot of being very ergonomic, optional GC, and great interop with C. I can drop down to assembly any time I want, or write a piece of C code to do something exotic, but the rest of the system is pure Nim. It's also quite fast.

[1] https://github.com/khaledh/axiom

I've used both to work on a hobby OS project (Nim[1], Zig[2]). I very much prefer Nim. Code is succinct, elegant, and lets you focus on your core logic rather than fighting the language.

Zig is nice and I like its optionals support and error handling approach. But I was put off by its noisy syntax, e.g. !?[]u8 to represent an error union of an optional pointer to a many-pointer of uint8. Also having to prepare and weave allocators throughout most of the code that needs to dynamically allocate (which is most of the code) gets in the way of the main logic. Even little things like string concatenation or formatting becomes a chore. Zig also doesn't have dynamic dispatch, which makes polymorphic code hard to write; you have to work around it through some form of duck typing. In the end I realized that Zig is not for me.

[1] https://github.com/khaledh/axiom

Niceness is subjective, but Nim is just as valid an addition to that group. Nim compiles to C and has had an --os=standalone mode for like 10 years from its git history, and as mentioned else-thread (https://news.ycombinator.com/item?id=36506087) can be used for Linux kernel modules. Multiple people have written "stub OSes" in it (https://github.com/dom96/nimkernel & further along https://github.com/khaledh/axiom).

While it can use clang as a backend, Nim does not rely upon LLVM support like Zig or Rust (pre-gcc-rust working). Use on embedded devices is fairly popular: https://forum.nim-lang.org/search?q=embedded (or web search).

Latency-wise, for a time, video game programming was a perceived "adoption niche" or maybe "hook" for Nim and games often have stringent frame rendering deadlines. If you are interested in video games, you might appreciate https://github.com/shish/rosettaboy which covers all but Ada in your list with Nim being fastest (on one CPU/version/compiler/etc). Note, however, that cross-PL comparisons are often done by those with much "porting energy" but limited familiarity with any but a few of the PLs. A better way to view it is that "Nim responds well to optimization effort" (like C/Ada/C++/Rust/Zig).

If anyone is interested, I have a couple of implementations of booting under UEFI and getting a bunch of info about the system (don't expect a functioning system, they just boot and dump some info):

Nim: https://github.com/khaledh/axiom

Zig: https://github.com/khaledh/axiom-zig (this one goes into depth in disassembling ACPI DSDT bytecode)

I wasn't ready to share it yet, but here it goes[1]. It's at a very early stage, but should give you an idea of how to get things up and running under Nim.

I didn't avoid malloc. I provided a simple bump pointer based heap to get things going. Later I'll have to separate things into a UEFI bootloader and a proper kernel image, each with its own allocator (the bootloader will use UEFI memory allocation services, and the kernel will have its own heap).

[1] https://github.com/khaledh/axiom

Couple times in the past I have implemented GPU-targeted GUI renderers, here’s an example: https://github.com/Const-me/Vrmac?tab=readme-ov-file#vector-... https://github.com/Const-me/Vrmac/blob/master/Vrmac/Draw/VAA...

2D graphics have very little in common with game engines. The problem is very different in many regards. In 2D, you generally have Bezier and other splines on input, large amount of overdraw, textures coming from users complicate VRAM memory management. OTOH, game engines are solving hard problem which are irrelevant to 2D renderers, like dynamic lighting, volumetric effects, and dynamic environment.

> Part of Panama

Most real-live C APIs are using function pointers and/or complicated data structures. Here’s couple real-life examples defined by Linux kernel developers who made V4L2 API: [0], [1] The first of them contains a union in C version, i.e. different structures are at the same memory addresses. Note C# delivers the level of usability similar to C or C++: we simply define structures, and access these fields. Not sure this is gonna be easy in Java even after all these proposals arrive.

For a managed runtime, unmanaged interop is a huge feature which affects all levels of the stack: type system in the language for value types, GC to be able to temporarily pin objects passed to native code (making copies is prohibitively slow for use cases like video processing), code generator to convert managed delegates to C function pointers and vice versa, error handling to automatically convert between exceptions and integer status codes at the API boundary, and more. Gonna be very hard to add into the existing language like Java.

> "Vector API" JEP

That API is not good. They don’t expose hardware instructions, instead they have invented some platform-agnostic API and implemented graceful degradation.

This means the applicability is likely to be limited to pure vertical operations processing FP32 or FP64 numbers. The rest of the SIMD instructions are too different between architectures. A simple example in C++ is [2], see [3] for the context. That example is trivial to port to modern C#, but impossible to port to Java even after the proposed changes. The key part of the implementation is psadbw instruction, which is very specific to SSE2/AVX2 and these vector APIs don’t have an equivalent. Apart from reduction, other problematic operations are shuffles, saturating integer math, and some memory access patterns (gathers in AVX2, transposed loads/stores on NEON).

> most of these are not done / not in a stable LTS Java release yet

BTW, SIMD intrinsics arrived to C# in 2019 (.NET Core 3.0 released in 2019), and unmanaged interop support is available since the very first 1.0 version.

[0] https://github.com/Const-me/Vrmac/blob/master/VrmacVideo/Lin...

[1] https://github.com/Const-me/Vrmac/blob/master/VrmacVideo/Lin...

[2] https://gist.github.com/Const-me/3ade77faad47f0fbb0538965ae7...

[3] https://news.ycombinator.com/item?id=36618344

I have minimal experience with Rust. OTOH, programming C++ for living since 2000, with a few gaps when I used other languages like Obj-C and C#.

I agree C++ is very hard to learn if you only have experience with higher-level languages like Python and Scala. I think there’re two reasons for that.

C++ is unsafe. There’s no way around this one, it was designed that way, like C or assembly. Still, with modern toolset it’s not terribly bad. Compilers print warnings, BTW I typically ask them to treat warnings as errors to deliberately fail the build. On Windows, a combination of debug build, debug C runtime, and visual studio debugger helps tremendously. Linux compilers have these sanitizers (address, memory, thread, undefined behavior) which are comparable, they too sacrifice runtime speed for diagnostics and debuggability.

Another reason, the language itself is very complicated, especially the templates. However, just because something is in the language doesn’t mean it’s a good idea to use it. You don’t need to be familiar with that stuff unless doing something very advanced, like customizing the Eigen C++ library. Don’t follow the patterns found in the standard library: unlike your code, that library has good reasons to use that template BS. If instead of templates you do something else, C++ becomes much easier to use, and most importantly other people will still be able to read and understand your code. Another reason to avoid excessive template metaprogramming, it slows down the compiler, because template-heavy code often needs to be in headers as opposed to cpp files.

P.S. If you don’t need extreme levels of performance (defined as “approach the numbers listed in CPU specs”, the numbers are FLOPS or memory bandwidth), and you don’t need the ecosystem too much, consider C# instead of C++. Much faster than Python, often faster than Scala or Java, easy integration with C should you need that (same as Rust, much easier than Python or Java), the only downside is these ~100MB of the runtime. The reputation is weird, but technically the language and runtime are pretty good. For example, here’s a C# library which re-implements a subset of ffmpeg and libavcodec C libraries: https://github.com/Const-me/Vrmac/tree/master/VrmacVideo

BTW, I did that too for 32-bit ARM Linux on Raspberry Pi 4, back in 2020: https://github.com/Const-me/Vrmac/tree/master/VrmacVideo Unlike Uno, my implementation doesn’t use libVLC and is written mostly in C#, only audio decoders are in C++. To decode video, I directly consume V4L2 Linux kernel APIs.

Doing that for decades.

An app for Windows phone, downloaded 140k times: https://github.com/Const-me/SkyFM

Cross-platform graphics library for .NET: https://github.com/Const-me/Vrmac

Recently, offline speech-to-text for Windows: https://github.com/Const-me/Whisper

At this point, I consider side projects like that as a hobby.

I think this needs much more complexity to be useful.

For the rendering, ideally it needs GPU support.

Input needs much more work, here's an overview for Windows: https://zserge.com/posts/fenster/

Windows' Sleep() function has default resolution 15.6ms, that's not enough for realtime rendering, and relatively hard to fix, ideally need a modern OS and a waitable timer created with high resolution flag.

Here's my attempt at making something similar, couple years ago: https://github.com/Const-me/Vrmac

I agree about Python or PHP.

However, for Java or modern C#, in my experience the performance is often fairly close. When using either of them, very often one doesn’t need C++ to be good enough.

Here’s an example, a video player library for Raspberry Pi4: https://github.com/Const-me/Vrmac/tree/master/VrmacVideo As written on that page, just a few things are in C++ (GLES integration, audio decoders, and couple SIMD utility functions), the majority of things are in C#.

To be fair, in modern GL versions they fixed some of these things. In GLES 3.1 which I used a lot on Pi4 https://github.com/Const-me/Vrmac/ GPU vertex buffers and shaders worked fine, GLSL compiler in the drivers worked fine too.

However, others issues are still present. There’s no shaders bytecode, they have an extension to grab compiled shaders from GPU driver to cache on disk, but it doesn’t work. The only way to create shaders is separate compile and link API calls. Texture loading and binding API is still less than ideal.