I'm reminded of a talk I enjoyed about "extreme rewriting" [1] — how rewriting the same code many times (in certain contexts) can help uncover powerful underlying abstractions.
It makes intuitive sense to me that this would be true in complex domains (e.g. legacy code) where you really need to find the right solution, even if it takes a bit longer. Our first ideas are rarely our best ideas, and it's easy to get too attached to your first solution and try to tweak it into shape when it would be better just to start fresh.
Comments
I'm reminded of a talk I enjoyed about "extreme rewriting" [1] — how rewriting the same code many times (in certain contexts) can help uncover powerful underlying abstractions.
It makes intuitive sense to me that this would be true in complex domains (e.g. legacy code) where you really need to find the right solution, even if it takes a bit longer. Our first ideas are rarely our best ideas, and it's easy to get too attached to your first solution and try to tweak it into shape when it would be better just to start fresh.
[1]: https://www.hytradboi.com/2025/03580e19-4646-4fba-91c3-17eab...