I think SharpLab works quite well for short pieces of code with as little boilerplate code as possible. One way to decrease the necessary boilerplate is to use Roslyn scripting mode (i.e. compiling with CSharpParseOptions.Kind set to Script). So, I think allowing that would be useful.
The resulting decompilation or disassembly would not be representative of regular C#, so it would be mostly useful with the other kinds of Results.
I have found https://github.com/ashmind/SharpLab/issues/4, which I think talks about this. As far as I know, Roslyn scripting is stable, just like the rest of the API in Roslyn NuGet packages.
I think SharpLab works quite well for short pieces of code with as little boilerplate code as possible. One way to decrease the necessary boilerplate is to use Roslyn scripting mode (i.e. compiling with
CSharpParseOptions.Kind
set toScript
). So, I think allowing that would be useful.The resulting decompilation or disassembly would not be representative of regular C#, so it would be mostly useful with the other kinds of Results.
I have found https://github.com/ashmind/SharpLab/issues/4, which I think talks about this. As far as I know, Roslyn scripting is stable, just like the rest of the API in Roslyn NuGet packages.