Gauthier Segay

@charlesroddie sure, what do you suggest to make it relevant? given @vzarytovskii feedback, I was tweaking it to > The following attributes 'NoEquality', 'ReferenceEquality', 'StructuralEquality', 'NoComparison' and 'StructuralComparison' are not...

This is the current state in this branch: The message only clarifies that the offending attribute is meant for DU & records, and it will only list the attributes employed,...

> Constructor-like initialization for properties instead of sequential assigning after creation @Lanayx, this could also apply to any method call that returns an object. Some of the analyzers will require...

Regarding custom settings, it comes to me that an analyzer referencing another one would possibly be able to export a chunk of settings. Say a blacklisted types analyzer would be...

> Your idea would be somewhat related to having the analyzers distributed as source (or source-like), so that the compiler could investigate them. The API surface analysis would work for...

https://github.com/fsharp/fslang-suggestions/issues/414 for `[]`. @baronfel, could you help editing the suggestion: * replacing places showing the attribute to `[]` * adjusting the title to "`[]` to require the use of named...

related: https://github.com/fsharp/fslang-suggestions/issues/806 & https://github.com/dotnet/fsharp/issues/1019#issuecomment-1819961451

What about the semantic coloring that takes care of highlighting `IDisposable` in a different manner? https://github.com/dotnet/fsharp/blob/99514c0fafa1f4a9ddf63e0439ec8804d87276eb/src/Compiler/Service/SemanticClassification.fs#L39 IIUC, we'd also like the same semantic classification to apply to `IAsyncDisposable`.

@oleksandr-bilyk I think right now it does already do it, if you pass the lambda in the call. Also, should this work with any delegate type instead? I'm not keen...

Likely a special case for #1131