[jeffadotio]

I really like this about Rust, implementing equality is just another trait. The same is true of comparison operators. If you don’t want a type to be able to be compared by == you can make it always evaluate to false or even panic and give a message that the type cannot be used in that way. That might aggravate coders who do not test comprehensively but it is an option.

Edit: This is a response to the concerns of the article. For general info on this it’s a great introduction to Rust’s awesome documentation.

https://doc.rust-lang.org/std/cmp/trait.Eq.html

[tus88]

You mean there is no compile time way to accomplish that? Doesn't sound very Rust-like.

[re]

It is determined at compile time; I'm not sure why GP is suggesting returning false if you don't want types to be compared. By default structs are not comparable. If you opt-in to the standard PartialEq implementation using the `derive` annotation[1], it will only allow a struct to be compared for equality with other structs of the same type. You can choose to add implementations to compare against other types[2].

[1] https://doc.rust-lang.org/book/ch05-02-example-structs.html#...

[2] https://doc.rust-lang.org/std/cmp/trait.PartialEq.html#how-c...

[artursapek]

Rust is such a delightful language.

[jeffadotio]

“Option” is the operative word. A type that does not implement the correct trait cannot be used in a way that requires the trait. If you are writing anything non-trivial that would probably be preferable. It requires nothing, that is the default behavior, so I did not bother to mention it.

[aliceryhl]

I don't know why they would make it always return false. In Rust you can just not implement Eq for the type, and it becomes a compile error to try and compare it.