Tony Morris

1. Stack causes more than zero problems. 2. Stack demonstrates no practical benefit. Ergo, Stack is a net penalty. It's not a trade-off. QED. You probably disagree with (2), in which case, demonstrate otherwise. No vagueness.

Stack - manages ghc versions for projects - provides automatic sandboxing with dependency sharing where appropriate - handles git dependencies - file watch is useful, even when ghcid is available

All of these were solved problems before stack existed. What is it do you think we were doing all those years? Waiting for this groundbreaking thing called stack?

I have no idea what you or anyone else was doing before stack, because the instructions given to beginners were fraught with problems that stack solved. As far as I know, cabal hell was a common experience.

You make this strange claim again, and yet I had been teaching FP for all those years, without all the scary problems with which someone was supposedly fraught. And it's not just me. What is it do you think we were doing when we were learning?

I have personally experienced big problems with cabal hell, and have helped dozens of people with similar problems. There absolutely was a problem that stack solved. Maybe the problem was using cabal at all? But that's how people said to use haskell.

Cabal hell never existed. You experienced a different problem. You called it cabal hell because that's what you were told to do. I find no other reasonable explanation. By "people", you mean, "the other beginners." I am sorry that happened to you.

What would you have called the problem that everyone else called cabal hell?