On Sun, 2021-09-05 at 08:15 +0200, 'Axel Wagner' via golang-nuts wrote: > Whether that factor is 2 or 1.25 doesn't matter. It's just a bit of > an optimization to make it 2 for small values - it doesn't terribly > matter what's "small" for this heuristic.
There is a reason for choosing a factor less than 2; released blocks during reallocation are able to be reused in later allocations (for the same slice when grown — something that cannot happen for the case of 2 or greater). Interestingly, Go uses the 1.25× factor while others appear to use values in the range of 1.5× to less than 2×, with a claim that phi is the optimum. There is no discussion in the introduction of append for the choice of the growth factor: https://codereview.appspot.com/2757042 -- You received this message because you are subscribed to the Google Groups "golang-nuts" group. To unsubscribe from this group and stop receiving emails from it, send an email to golang-nuts+unsubscr...@googlegroups.com. To view this discussion on the web visit https://groups.google.com/d/msgid/golang-nuts/bcf24b8e53f3bc6aafb186ced57b614b4ecf4891.camel%40kortschak.io.
