move FCS to netstandard2.0 and purge old net461 related code

[baronfel]

This moves FCS to NS2.0 only.

The ProjectCracker, ProjectCrackerTool, and MsBuild.v12 projects have been removed, as those were net461 only. In addition, the FscExe and FsiExe samples have been removed, as they relied on .Net Framework Remoting, which has no nice analog in a .Net Standard/.Net Core world.

The netfx.props were removed as well, because nothing else was using them.

[auduchinok]

We currently still target net461 and there might be some issues with netstandard2.0 assemblies according to discussion with @cartermp.

[baronfel]

I wondered about that. If the timing's not right that's cool, this took like half an hour so it'd be easy to do later.

[cartermp]

@auduchinok If you wouldn't mind testing a nupkg of this to see if everything works that'd be good. There may be runtime errors for the net461 target but there may not; unfortunately there's no way to analyze that

[auduchinok]

Rider 2020.1 EAP2 uses the netstandard2.0 FCS build. We're still looking if there are any problems with it and if something bad is found or reported I'll let you know.

[cartermp]

Thanks @auduchinok - tentatively, let's see if we can take this in once that build ends up shipping outside of EAP.

[cartermp]

linking #8722

[cartermp]

Unfortunately the codebase has diverged a lot. We want to take this in, but only once we can confirm it doesn't hose Rider will we continue with it. I'll close this out for now, since picking this up again will involve merging conflicts that could end up being about the same amount of time spent doing this in the first place.

[auduchinok]

We've been using the netstandard2.0 build for some time now and with the dependency manager fixes it currently works fine. There are some issues when running on Mono but it's likely we won't be using Mono in 2020.2 even as a fallback runtime anymore.
I'm going to grab a newer FCS build soon to test it all again, but now it seems it's more or less safe to go netstandard2.0 FCS for us.

[cartermp]

will close this out in favor of #10135 on account of everything being different now