4 ms·
Yes. I agree. I thought the essay might be useful or informative, but about a third of the way through I realized that this was one of these "look how clever I
by DanielBMarkham 6y ago
Yes. I agree. I thought the essay might be useful or informative, but about a third of the way through I realized that this was one of these "look how clever I am" essays.
This would have been a good launching-off point to talk about the linker, or language design, which we did a bit of towards the end but in the wrong context, in my opinion. Instead this author was a bit too clever by half; a shorter article more focused on practical limitations might have been much more helpful to coders.
- Animats 6y ago> "look how clever I am" essays. Yes. If you had general overload on return type, you could write big nested expressions and force the compiler to examine all the overload possibilities. With enough effort, you could write optimizers for functional programming that way, ones that figured out when you can create a pipeline and avoid storing an intermediate result. Plus combinatoric explosion in compile times. And then you would need more advanced algorithms to find a path through the type maze efficiently. What fun! Good for at least three PhD theses. Don't go there. Someone will have to maintain that code.