I'm using mpz_prevprime again this week and I'm so happy it's present as part of the default libgmp-dev on my distro. Thanks for being open to me adding this years ago and for everyone's help getting the code to the place it is!
On Sat, Nov 28, 2020 at 9:08 AM Marco Bodrato <bodr...@mail.dm.unipi.it> wrote: > Ciao, > > Il 2020-11-16 21:23 Seth Troisi ha scritto: > > Thanks Marco! I'm really happy this is merged. > > Your code was merged with a storm of new code that has been reversed. > But now it's in again! > > > Was there any specific things you thought could be improved? Happy to > > look at those too. > > Memory usage, in my opinion, is not optimal. > > The current code generate a primesieve, to just loop on it once for > building a list of gaps, that is usually used just once... I'm still not > completely convinced that this structure is the best we can do. > > In the meanwhile I wander if a further modularization could allow us to > build a type to store the "current state" and functions to sequentially > find the next primes... > > Ĝis, > m > > _______________________________________________ gmp-devel mailing list gmp-devel@gmplib.org https://gmplib.org/mailman/listinfo/gmp-devel