--- In primenumbers@y..., "Andrey Kulsha" <Andrey_601@t...> wrote:
> > I cannot address your feature request, but generally you
> want to check
> > only even b values.
> GFNSieve goes from bmin with step 2 to bmax, so if bmin is
> odd, sieving is senseless.
> I got used that Proth automatically checks min, max and
> step. I've entered bmin=1 and now I see: all sieved bases
> are odd. :-)
This is a logic bug, and will be fixed shortly. The engine used to
check for and fix this condition, so I am not sure what is not
working. It is a very trival fix. I may have inadvertently removed
some logic checking when I added the logic checking to make sure that
the b's were not "too" large. I use the benchmarking numbers from
Proth as the largest workable size for GFNSieve, and if larger numbers
are entered, the program will set them to be in this "valid" range.
> Neither GFNSieve, nor Proth check this. So we can see
> something like this:
> 866761^512 + 1 ? Computing power [0.00%]
> * * *
> Thanks to David Underbakke for GFN links.
> Best wishes,