Replies: 3 comments 1 reply
|
You can contribute a C# implementation if this is your use case. |
|
yeppi, language wars, I always love those :) seriously though, .net / java / python / go / javascript all require "runtime engine" - a large chunk of code specific to that language and required to run it. C / C++ / Rust do not - they can be (somewhat) easily used directly by all of the above. There is reason why every "higher-level" (runtime) languages have C-style bindings, and also why 1/3 of published python packages are in Rust - it offers a path to have extremely fast and portable shared code in a system language. So if we build system-level encoder and decoder, it paves a way to reuse. If we build something in C# (which is a great language btw - I spent almost 20 years with it), it will only serve the .net MLT-using community. Which obviously brings the question - how big is that community - who will be the users of that lib? And if the answer is uncertain, it might make much more sense to have good bindings to the Rust implementation rather than fully C# one. A quick search showed this interesting project for that - https://github.com/Cysharp/csbindgen/ |
I did not know that you could use Rust in C#. I understand the interoperability of C/C++/Rust to higher-level runtime languages now. But then why do you have a Java implementation? I see the only way to interop Rust -> Java is to use Foreign Function Interface (FFI). The same way you would if you created a C# implementation. |
Uh oh!
There was an error while loading. Please reload this page.
The .NET Core framework has done some amazing optimization work. The framework is not only mature but still pushing boundaries and massively growing in popularity. Its recently surpassed Javascript in the Tiobe index rankings. Rust is way below C# at 13th. I don't understand the reasoning why we don't have a .NET implementation?
All reactions