I would disagree. The purpose of big O in the STL is not to allow you to compare different algorithms, but allow you to reason about how that will scale with input size. If I test perf with a size of 1000 and then it’s declared to be linear, I have a pretty good idea if that’s the right tool for the job.
Do you have any examples of “better” algorithms that the STL is not able to use? Another idea in STL is to provide many algorithms, which is why there are 3 sorts.
Do you have any examples of “better” algorithms that the STL is not able to use? Another idea in STL is to provide many algorithms, which is why there are 3 sorts.