On 12 October 2016 at 06:58, Elliot Gorokhovsky email@example.com wrote:
So I got excited here. And the reason why is that I got those numbers *on Tim's benchmark*. When I got these kinds of numbers on my benchmarks, I figured there was probably a problem with they way I was timing, and certainly the gains couldn't be as extreme as they suggested. But this is on a benchmark that's already in the codebase!
Thanks for the clearer write-up - this is indeed very cool, and it's wonderful to see that the new assumptions permitted by Python 3 getting stricter about cross-type ordering comparisons may lead to speed-ups for certain common kinds of operations (i.e. sorting lists where the sorting keys are builtin immutable types).
Once you get to the point of being able to do performance mentions on a CPython build with a modified list.sort() implementation, you'll want to take a look at the modern benchmark suite in https://github.com/python/performance