A few times, but it's not something i insist much on, i tend to see the opposite problem, rather.
Significantly
I try to clarify on what "premature" means, that is, not going for the dumbest solution that could possible work, no matter how inefficient it is, but not going overboard about implementing overly complex ones until you get a real sense of where the bottlenecks are going to be, usually using profiling, but also still do design so your solution scale to the need.
Add comment