[sage-devel] Re: Policy regarding timing examples/benchmarks in the docs

2014-12-19 Thread Simon King
Hi Nathann, On 2014-12-18, Nathann Cohen wrote: > So something like: > - Big speedup on all architectures --> timings > - Comparable results --> a big red warning or a note saying "try for > yourself, we don't know your hardware" That sounds like a reasonable rule of thumb. Here, it is like th

[sage-devel] Re: Policy regarding timing examples/benchmarks in the docs

2014-12-18 Thread Nathann Cohen
Yo ! I would say that it would be nice to have timings if the timings say that "for this computation, one is clearly better than the other". So that the actual figures do not matter as long as there is a big speedup in some direction. If the timings are comparable and depend on the architectur