Peter Dimov-2 wrote
We determine whether the library works by testing it, and we determine the library's performance by timing it. We're empiricists.
Of course, if Steven wants to provide formal proofs of correctness and complexity requirements, that would be very nice of him; I just don't think it's fair to demand it.
+1 In our context, testing trumps proof. This is because all such "proofs" count operations of some sort. And operations (e.g. compare, relink, dereference, etc) vary between algorithms. Also I agree that we should stick to our our stated standards. Robert Ramey PS - I'm was sort of disappointed that the Postman's sort wasn't referred to. It was published many years ago and has been cited numerous times. It has several times won a contest held by Microsoft Research for the world's fastest sort. To prove it, I have two medals hanging on my door knob (each with the name spelled wrong - in different ways). Robert Ramey -- View this message in context: http://boost.2283326.n4.nabble.com/review-Formal-review-period-for-Sort-libr... Sent from the Boost - Dev mailing list archive at Nabble.com.