spec VS Oberon

You can parse many things from this file, what are you trying to extract?

https://github.com/WebAssembly/spec/blob/main/document/core/...

Still strikes me as super sad JS never got SIMD support. It seemed like there were some strong candidate specs. On Node there are some add-on npm libraries that implement.

My understanding was the main protest was that we would get wasm & some certain implementers said they wanted to focus their energy on wasm.

That was well over half a decade ago & wasm is still in incredible infancy, with basically only statically linked capabilities in the spec.

Wasm SIMD proposal itself only merged into wasm in November 2021. https://github.com/WebAssembly/spec/pull/1391

It seems really unfortunate to have decided to keep JS the slow inferior language.

Blazor and WASM itself (outside of dotnet) are relatively new tools and they already show impressive results. They will keep getting better with every release. E.g. this proposal https://github.com/WebAssembly/spec/blob/main/proposals/simd/SIMD.md which should bring WASM closer to "near native speed". Blazor already started working on it true.

Big fan of this author's work.

They have a Gameboy emulator written in C, which can be compiled to WASM and run in the browser.

https://github.com/binji/binjgb

I learned a lot from the code.

Also I love this project with a bunch of demos in hand-written WebAssembly Text (WAT) format, which is like low-level Lisp that works only with raw memory, numbers, and minimal syntax.

https://github.com/binji/raw-wasm

Then I discovered the same author is quite active in the WebAssembly ecosystem, including specs and tooling. Fascinating stuff!

https://github.com/WebAssembly/spec

https://github.com/WebAssembly/wabt

[The WebAssembly specification (https://webassembly.github.io/spec/) maintains that the standards apply to more than just the browser host, but also to any other compliant host runtime (what the specification refers to as an embedder).

Heya,

(1) Thank you for implementing this in JSC!! I hope they take it, it makes it into Safari, and the tail-call proposal advances.

(2) I don't think you are exactly right about the call stack being observable via thrown exceptions. There's no formal spec for the v3 exceptions proposal yet, but in the documents and tests, there's nothing that would change in WebAssembly core to make the call stack observable. It's true that the proposal amends the JS API (but only the JS API) to describe a traceStack=true option; from Wasm's perspective I understand that's just an ordinary exception that happens to include an externref value (just like any other value) to which Wasm itself attaches no special significance. The engine can attach a stack trace if it wants, but there's no requirement (here) about what that stack trace contains or whether some frames might have been optimized out.

(3) I think the real reason that a Wasm engine can't implicitly make tail calls proper is that the spec tests forbid it, basically because they didn't want the implementation base to split by having some engines perform an optimization that changes the space complexity of a program, which some programs would have started to depend on (the spec tests say: "Implementations are required to have every call consume some abstract resource towards exhausting some abstract finite limit, such that infinitely recursive test cases reliably trap in finite time. This is because otherwise applications could come to depend on it on those implementations and be incompatible with implementations that don't do it (or don't do it under the same circumstances.)" More discussion here: https://github.com/WebAssembly/spec/issues/150

The WASM paper discusses that in the final section: https://github.com/WebAssembly/spec/blob/master/papers/pldi2017.pdf

WebAssembly has a formal specification that includes both operational semantics and natural language-based descriptions of everything in the language. The official repository also has a lot of tests. Besides tests, you should be able to find lots of examples by searching.

> Sure there's a small overhead to smart pointers

Not so small, and it has the potential to significantly speed down an application when not used wisely. Here are e.g. some measurements where the programmer used C++11 and did everything with smart pointers: https://github.com/smarr/are-we-fast-yet/issues/80#issuecomm.... There was a speed down between factor 2 and 10 compared with the C++98 implementation. Also remember that smart pointers create memory leaks when used with circular references, and there is an additional memory allocation involved with each smart pointer.

> Garbage collection has an overhead too of course

The Boehm GC is surprisingly efficient. See e.g. these measurements: https://github.com/rochus-keller/Oberon/blob/master/testcase.... The same benchmark suite as above is compared with different versions of Mono (using the generational GC) and the C code (using Boehm GC) generated with my Oberon compiler. The latter only is 20% slower than the native C++98 version, and still twice as fast as Mono 5.

Great, thanks!

There are books online for free, e.g.

https://people.inf.ethz.ch/wirth/ProgInOberonWR.pdf

and https://ssw.jku.at/Research/Books/Oberon2.pdf

Oberon+ is a superset of Oberon 90 and Oberon-2. Here is more information: https://oberon-lang.github.io/, and here is the current language specification: https://github.com/oberon-lang/specification/blob/master/The.... I already had valuable feedback here on HN concerning the channel extensions. Further research brought me to the conclusion, that Oberon+ should support both, channels and also monitors, because even in Go, the sync package primitives are used twice as much as channels. Mutexes and condition variables can be emulated with channels (I tried my luck here: https://www.quora.com/How-can-we-emulate-mutexes-and-conditi...), but for efficiency reasons I think monitors should be directly supported in the language as well, even if it might collide with the goal of simplicity.

Feel free to comment here or e.g. in https://github.com/rochus-keller/Oberon/discussions/45.

> Does anyone know why Wirth never modernized his style?

Readability. It's easier to read the source code with uppercase keywords. (I think Wirth once said that code is written once but read many times). See this source code - https://raw.githubusercontent.com/rochus-keller/OberonSystem... - to get an idea of this (the uppercase keywords allow you to easily scan the blocks of code). Ofcourse, one can claim that the same can be achieved better today with colour-coded keywords.

If I remember right, the Oberon+ IDE - https://github.com/rochus-keller/Oberon - gives you an option to disable this and use lowercase keywords.

> gain a deep understanding of it .. generate smaller subsets of the system

You can use the OberonViewer for this purpose with the original source code, or the Oberon IDE with a version of the Project Oberon System which runs with SDL on all platforms, see https://github.com/rochus-keller/oberon/#binary-versions and https://github.com/rochus-keller/OberonSystem/tree/FFI

> Regardless, which one is more likely to be ported to a different architecture in the future?

Not sure I understand the question. I'm talking about CPU architectures. The current implementation is in x86 assembler. So if you want to run it on AMD64 or ARM, then you have to replace all assembler files, in the present case probable the full source code.

> what are the comparative performance benchmarks of the low-level language versus the high-level language?

I don't have any measurements. But consider that many operating systems are implemented in C (e.g. Linux) with only isolated parts in assembler, so it is easier to port to other architectures. Linux apparently is fast enough and available for nearly every CPU. Oberon in contrast to C is garbage collected, which also affects performance. I have measurements comparing the same benchmark suite implemented in C++ and in Oberon, where the former is about 22% faster (see https://github.com/rochus-keller/Oberon/blob/master/testcase...).

I actually had a similar problem some years ago and finally moved away from LLVM because of complexity, continuous research effort and performance. My current Oberon+ implementation works like this: the CIL code generator together with Mono is used during development, integrated with the IDE, using the debugging features integrated in Mono; to deploy the application and to gain another factor 2 of performance C99 instead of CIL can be generated and compiled with any compatible toolchain. Here are some performance measurements: https://github.com/rochus-keller/Oberon/blob/master/testcases/Are-we-fast-yet/Are-we-fast-yet_results_linux.pdf. Compiling to CIL is very fast and the time Mono needs to compile and run is barely noticable.

> annoying aspects was requiring the .NET runtime ... OpenJDK is a blessed implementation in a way that Mono never was

Which is unjustified, because Mono CLR is just a single executable less than 5 MB which you can download and run without a complicated installation process (see e.g. https://github.com/rochus-keller/Oberon/#binary-versions ). AOT compilation on the other hand is a huge and complex installation depending on a lot of stuff including LLVM, and the resulting executables are not really smaller than the CLR + mscorlib + app.