-
When defining an unboxed record, one gets to define new instances for type classes on the Agda side. Those instances never get compiled, and code that uses instances over boxed values will compile to …
flupe updated
1 month ago
-
something like
```agda
bad : (Bool → Bool) → Bool
bad = λ x → {!x x!}
```
results in "Cannot refine functions with 10 or more arguments" message instead of giving a proper reason that functio…
-
Since this is a book _for functional programmers_, we ought to make liberal use of lambda expressions. But the Preliminaries section should introduce explicitly what our notation is (i.e. do we use th…
-
Built-in sorts are highlighted as strings, but shouldn't be:
I suggest using a scope like `support.type.sort.agda` instead.
-
> We've actually gone out of our way in agda-categories to make as many of these things hold judgementally. For `1 -> C` vs `C`, that is hopeless, as they are fundamentally different even if equivale…
-
> Revisiting #1544, I am (now) conscious that the *opposite* choice is already made regarding the name of the homomorphism property in `Algebra.Morphism.Structures.IsMagmaHomomorphism`, which for what…
-
Hello,
First, let me thank you for writing and maintaining this extension: agda would be very difficult without it.
I've noticed that system calls via the Reflection.External interface fail if t…
-
Maybe now that `opaque` is implemented in Agda 2.6.4 we should revisit our solution to https://github.com/agda/agda-stdlib/issues/1753.
This is exactly the situation that the new `opaque` mechanis…
-
Hi!
Plugins like https://github.com/tpope/vim-commentary check the `commentstring` to get the formatting required to comment out lines. The other plugin I used had it https://github.com/derekelkin…
siers updated
2 years ago
-
It would be good to have some kind of testing that ensures our implementation of the Haskell prelude is correct. For example, for each function we could compile our implementation to Haskell and use Q…