[Proposal]: Edit speclets to distinguish features already delivered from potential future enhancements

[BillWagner]

We've been publishing the feature speclets on Learn so that customers can easily access the detailed information about new features.

There have been a few comments from readers that are confused because the speclets include specification language for features that haven't been implemented yet. For example:

• Low level struct improvements The section on scoped has been implemented. Much of the remaining sections could be implemented later
• Covariant return types has a section on explicit interface implementation that hasn't been implemented.

There are likely others.

One fix would be to make the following edits:

1. Separate out sections that haven't been implemented into new proposals in the Proposals.
2. Link to the new "proposed additions" speclet from the existing specs to provide a roadmap for future work. That link would be an absolute GitHub link, so that readers on Learn would be redirected to GitHub. That would be a signal that it's a future enhancement.

That would benefit Learn readers, and provide a clear distinction for designers and the compiler team on what behavior has been implemented, and what behavior is still in a proposed state.

LDM should weigh in on this before editing the docs.

[333fred]

Another part to consider is the sections of the specs that call out invalid scenarios: I've had several questions on some of the DIMs examples of what doesn't work from people just copy/pasting examples.

[KathleenDollard]

@333fred We might want a docs feature on this. I'd love to see some different visual appearance when it is an example of what not to do, either because it won't work or it's just a terrible idea.

That is probably a docs infrastructure issue. @BillWagner should we open one?

[jcouv]

Found another one: the inline-arrays spec suggests that inline array types can be initialized from a collection expression, but that's not supported.
https://github.com/dotnet/csharplang/blob/main/proposals/csharp-12.0/inline-arrays.md#collection-literals

FYI @AlekseyTs @cston

[jnm2]

From the perspective of someone who has consumed the C# spec proposals as a library and tooling author, this spec documentation is well worth solving and delivering along with each release of C#.

My ideal would be to keep the long-term proposal separate from the shipped version of the proposal, moving stuff over as it is implemented, rather than requiring people to skip over sections.