Steve Dunham

This quick and dirty patch fixes it, but will probably break other backends. It is just to test if freeing the pointer works. We should check if it is also...

Racket version 8.4 [cs] has the same issue. I suspect every scheme has the issue since they can't know the difference between `prim__getEnv` (should never be freed) and `prim__readLine` (should...

I was thinking of filing this after a discussion on discord about a pack install failure. This occurs if you have `IDRIS2_INC_CGS=chez` set in the environment. I haven't dug deeper...

You probably don't want it set. It can be useful when working on Idris itself, but the tradeoff is that the compiler builds faster, but the resulting program runs slower....

I had this issue on OSX and eventually figured out the problem (for me) was uMatrix. The url in the figwheel output was http://0.0.0.0:3449 and then uMatrix blocked an XHR...

Some of the issues that Syzygies struggled with are fixed post 0.5.1. It's been a year and a lot of stuff has been fixed since then. Perhaps there should be...

I hit a similar error today when trying to write a combinator parser. (No holes involved.) What was happening for me is that there were recursive dependencies between lazy values....

As another example, this will error in scheme with `Exception in number->string: # is not a number` and overflow the stack on node: ```idris bar : List Int foo :...

This is also an issue if the `n` is not erased: ```idris data T : Nat -> Type where A : T Z B : (n : Nat) -> T...

Do you have an example of the behavior you want? I believe happens with this patch should be exactly the same behavior as mutual blocks. (The code for desugaring the...