I've been a huge proponent of file-scoped namespaces and it's in C# 10. I think it'll improve code readability and maintainability significantly despite being a seemingly minor change. It's also in line with "minimal API" and "top-level code" changes. I wrote about it here: https://ssg.dev/a-pledge-for-namespace-directives-in-c-15a32...

Discriminated unions is one of those features that everyone talks about, but never gets championed. Wish C# 10 already had this feature. https://github.com/dotnet/csharplang/discussions/2962

you can actually do them pretty easily in C# without language support anyways.

for instance, the F# example from https://fsharpforfunandprofit.com/posts/designing-for-correc...

can be done in C# like https://gist.github.com/keithn/ceeeed5f7eb567e1b2333747065d1...

almost similar amount of code, however, no exhaustive checking. Not that I find that to be much of a problem, but it would be nice

But who needs sum types when you have the visitor pattern :troll_face:

https://blog.ploeh.dk/2018/06/25/visitor-as-a-sum-type/

wow.....that is one way to do it.....yes. I would hate to think what a system would look like with many sum types in it

Maybe try F#? I know it isn't always an option but if you can, especially recently, its been quite productive to work in.

Static interface members enabling generic math and other cool features slipped the C# 10 quality bar. That is why the release looks a bit thin.

But honestly, with file scoped namespaces and global usings ... there is enough to see.

C# is already evolving fast enough that quite a few devs are uncomfortable with it. Speaking strictly for myself, I love getting new language features faster - but there's a balance to strike here for the overall good of the ecosystem. You don't want to drive out all the people who can't keep up.

I love C#10 for a few features. The language is developing rapidly. And C# is already productive. I would like to see fewer features in each release.