On 9/21/16 4:16 AM, Jon Degenhardt wrote:
Timing comparison of sort and topN, times in milliseconds:sort topN Field 2: 289 1756 Field 3: 285 148793 Field 4: 273 668906 The above times are for LDC 1.1.0-beta2 (DMD 2.071.1). Similar behavior is seen for DMD 2.071.2. This makes topN pretty much unusable.
I have it on my list to move https://arxiv.org/abs/1606.00484 into Phobos. Thanks for the data! -- Andrei