On Fri, 26 Feb 2016 07:54:26 +0100
isabella parakiss <izaberina_AT_gmail.com> wrote:
Hey izabera,
> the problem with factor is that any naive implementation will pale against
> the one in gnu coreutils.
>
> (gnu)
> $ time factor 1267650600228402790082356974917
> 1267650600228402790082356974917: 1125899906842679 1125899906842723
> real: 0m1.576s, user: 0m1.570s, sys: 0m0.003s
>
> (yours with gmp)
> $ time ./factor 1267650600228402790082356974917
> 1267650600228402790082356974917: 1125899906842723 1125899906842679
> real: 1m11.109s, user: 1m11.013s, sys: 0m0.013s
>
> (yours with tommath)
> $ time ./factor 1267650600228402790082356974917
> ^C interrupted after 20 minutes
>
> from at least 50x to more than 1000x slower than the gnu version.
> does this suck less?
Just use matlab or wolfram alpha. I see no need for such a tool in
sbase. If we lool at the history of factor(1), some guy came up with
it and through the years, there was no reason not to include it in the
package, however, the use for this software is very low.
Maybe we should go back to fixing real problems and improve tools that
people actually use.
Cheers
FRIGN
--
FRIGN <dev_AT_frign.de>
Received on Fri Feb 26 2016 - 09:04:23 CET