You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
asdf::Choreo '["a", "b"] (CLIm) Int
asdf =do x <- a `_locally` getInput "Enter a number"
x' <- (a `introAnd` a, x) ~> a @@ b @@ nobody
naked refl x'
I think I've written that so it's totally valid, but what happens if we try to project it to "c"? x' will be Empty, so naked can't work.
Since we need to determine at Haskell-Runtime which party we're projecting to, I don't think we can use our existing GDP approach on epp; i don't yet know how to fix this.
The text was updated successfully, but these errors were encountered:
My current idea is to re-write EPP to take a Member p census arg instead of a LocTm. For this to happen elegantly, we should first refactor Located so it can have an instance of TestEquality for Member. (although... having such an instance will require deciding how to handle a census that contains the same party twice.)
Then we can recover the old behavior with some kind of forLocs wrapper maybe?
Consider the choreography
I think I've written that so it's totally valid, but what happens if we try to project it to
"c"
?x'
will beEmpty
, sonaked
can't work.Since we need to determine at Haskell-Runtime which party we're projecting to, I don't think we can use our existing GDP approach on epp; i don't yet know how to fix this.
The text was updated successfully, but these errors were encountered: