5 ms·

Why not just use a stack data structure instead of using the call stack as a stack data structure? Each stack frame is going to take up a lot more space than a

• Jtsummers 1y ago
  You may disagree with their take on it, but they do address that in the write-up. > This approach works for simple cases but becomes extremely complex or impossible when any of these conditions apply: > 1. The algorithm transforms data structures rather than just evaluating them (e.g., optimizing an AST) > 2. Multiple recursive calls need to be coordinated (e.g., tree balancing algorithms) > 3. The algorithm doesn’t fit the tail-recursion pattern I would disagree with "impossible" (pretty sure it's never actually impossible, but some type systems and language features may work together to make it so, I suppose), but definitely agree with "extremely complex".
  • CyberDildonics 1y ago
    I definitely do disagree because if you distill it down to making a single recursive call or pushing a value on a stack with local scope, I would say pushing a value on the stack in simpler every time. If there are multiple recursive calls you could use multiple stacks instead. All of the debugging is going to be more difficult if you have to move through multiple call stacks to get at the previous values, where you could just see the entire stack values in a debugger or print out the stacks. For number 3, I think that doesn't apply, since tail-recursion is really a funky way of looping and not a funky way of using the call stack for a stack.