RecordValueAnalyser VS PolySharp

I've written a small code analyser for Roslyn, which works fine when included as a csproj. The source is at https://github.com/lookbusy1344/RecordValueAnalyser

Great improvements.

There is PolySharp project that enables you to use most of C#11 features in legacy .NET Framework: https://github.com/Sergio0694/PolySharp - Seems that C#12 features are planned to be implemented: https://github.com/Sergio0694/PolySharp/issues/78

I'm using PolySharp where I'm stuck with .NET Framework 4.6 and I don't have any issues.

Hope one day I'd see concise syntax for catch and/or try expressions: https://github.com/dotnet/csharplang/discussions/2734 - but there is a lot of resistance.

You can go one step further and add https://github.com/Sergio0694/PolySharp to your project and you can use new C# 8+ features that don't require the new runtime (like file scoped namesapces, nullable annotations etc).

Is there a reason you're trying to reinvent PolySharp?

You can still use C# 11 (with PolySharp), and you can still get a huge amount of new APIs via the various packages on NuGet that target .NET Standard 2.0. Plus you just get the usual .NET type system and BCL, which are still very good even if when using a framework from a few years ago.

Or the PolySharp NuGet package, which uses source generators to polyfill the language features you're actually using.

Here's some: - MVVM Toolkit - PolySharp - ComputeSharp

Well using PolySharp, of course! 😄

⚠️ Because PolySharp uses source generators, it doesn't work with the package.config file as stated in this issue. The issue says we need to use the SDK style .csproj, but just changing from package.config to Package Reference worked for me.

Some features like records and range require additional boilerplate classes to be added to become usable. You can DIY that or use PolySharp. Some features like static abstract members, default interface methods, etc require runtime support and can't be used.