Re:The Next Step for Generics
by alkonaut on Wednesday June 17, 2020 @03:56AM
(#23548950)
For me that type golfing is how I explore the problem. A lot of times you realize things about the problem that you didn’t understand before.
This is especially true with Rust where you also need to consider who owns what data. With C# it’s much too to model things that make sense as a type zoo but where the more important problems like ownership and access patterns is lost in the cracks.
The typical sign in Rust is when you need a lifetime on a struct so you realize it has someone else’s data. That usually means I stop to think before progressing. In C# it’s very easy to miss that big red flag.
Obviously it’s always worth revisiting and simplifying a design, but I still don’t think the initial design or designs were a waste of time if it gave me an understanding of the problem space.
This is especially true with Rust where you also need to consider who owns what data. With C# it’s much too to model things that make sense as a type zoo but where the more important problems like ownership and access patterns is lost in the cracks.
The typical sign in Rust is when you need a lifetime on a struct so you realize it has someone else’s data. That usually means I stop to think before progressing. In C# it’s very easy to miss that big red flag.
Obviously it’s always worth revisiting and simplifying a design, but I still don’t think the initial design or designs were a waste of time if it gave me an understanding of the problem space.