Zulip Chat Archive

Stream: Is there code for X?

Topic: x + y ⋖ x + z ↔ y ⋖ z


Violeta Hernández (Oct 28 2024 at 15:39):

Do we have some variant of this? If not, what are the necessary typeclass hypotheses to prove it?

Violeta Hernández (Oct 28 2024 at 15:39):

@Yaël Dillies

Yaël Dillies (Oct 28 2024 at 15:41):

No, we don't have it. The precise condition you need is that (x + ·) is an order embedding. It happens that we don't have a nice set of typeclasses that implies that

Yaël Dillies (Oct 28 2024 at 15:42):

CovariantClass α α (· + ·) (· ≤ ·) + ContravariantClass α α (· + ·) (· ≤ ·) should do

Violeta Hernández (Oct 28 2024 at 15:43):

Ehm, do you know what those typeclasses are named now? A recent refactor renamed them

Daniel Weber (Oct 28 2024 at 15:45):

There's still docs#CovariantClass and docs#ContravariantClass, but I think you can use docs#AddLeftMono and docs#AddLeftReflectLE

Yaël Dillies (Oct 28 2024 at 15:45):

See #13154


Last updated: May 02 2025 at 03:31 UTC