Remove POrd instance of PValue - add PPartialOrd
#528
Conversation
Add `PPartialOrd` instance instead
Plutarch/Bool.hs
Outdated
| @@ -89,6 +90,18 @@ class PEq t => POrd t where | |||
| infix 4 #<= | |||
| infix 4 #< | |||
|
|
|||
| -- | Partial ordering relation. | |||
| class PPartialOrd t where | |||
| pleq :: Term s t -> Term s t -> Term s PBool | |||
There was a problem hiding this comment.
should these methods be Plutarch level functions instead?
Plutarch/Bool.hs
Outdated
| @@ -89,6 +90,18 @@ class PEq t => POrd t where | |||
| infix 4 #<= | |||
| infix 4 #< | |||
|
|
|||
| -- | Partial ordering relation. | |||
| class PPartialOrd t where | |||
There was a problem hiding this comment.
Should this become a superclass of POrd? It'd require adding all the instances everywhere else
|
One possibility is to choose the design Rust chose, but when would you depend on a PPartialOrd constraint rather than POrd? We could move #< etc. to PPartialOrd, make a superclass relation, and then derive PPartialOrd for everything that has POrd, but would you ever use the PPartialOrd instance? Is it confusing that you can use < for some things, but only #< for others?
I feel like Rust's approach might be the best solution, but really, we want PartialOrd too.
|
|
Wouldn't Rust's approach be suboptimal in UPLC, if we adopt I imagine there won't be much user-written polymorphic code; when writing a contract people tend to know what type they're dealing with. It's only the library writers who will have to decide between |
|
I also think moving things into |
|
Moved the |
|
@L-as wait we still needed to add laws? |
|
@TotallyNotChase Well we can do that in another PR and it's not essential for 1.2 IMO. |
Closes #526