Re: [dev] [sbase][PATCH] Add factor(1)

From: Dimitris Papastamos <sin_AT_2f30.org>
Date: Fri, 26 Feb 2016 11:55:12 +0000

On Fri, Feb 26, 2016 at 12:45:27PM +0100, Mattias Andrée wrote:
> On Fri, 26 Feb 2016 11:39:07 +0000
> Dimitris Papastamos <sin_AT_2f30.org> wrote:
>
> > On Fri, Feb 26, 2016 at 12:33:37PM +0100, Mattias Andrée
> > wrote:
> > > On Fri, 26 Feb 2016 11:31:31 +0000
> > > Dimitris Papastamos <sin_AT_2f30.org> wrote:
> > >
> > > > On Fri, Feb 26, 2016 at 12:28:32PM +0100, Mattias
> > > > Andrée wrote:
> > > > > On Fri, 26 Feb 2016 11:26:15 +0000
> > > > > Dimitris Papastamos <sin_AT_2f30.org> wrote:
> > > > >
> > > > > > On Fri, Feb 26, 2016 at 12:19:32PM +0100, Mattias
> > > > > > Andrée wrote:
> > > > > > > On Fri, 26 Feb 2016 10:07:31 +0000
> > > > > > > Dimitris Papastamos <sin_AT_2f30.org> wrote:
> > > > > > >
> > > > > > > > On Fri, Feb 26, 2016 at 09:47:14AM +0100,
> > > > > > > > Mattias Andrée wrote:
> > > > > > > > > On Fri, 26 Feb 2016 09:43:02 +0100
> > > > > > > > > FRIGN <dev_AT_frign.de> wrote:
> > > > > > > > >
> > > > > > > > > > On Fri, 26 Feb 2016 09:37:12 +0100
> > > > > > > > > > Mattias Andrée <maandree_AT_kth.se> wrote:
> > > > > > > > > >
> > > > > > > > > > Hey Mattias,
> > > > > > > > > >
> > > > > > > > > > > Mostly random things, but regularly
> > > > > > > > > > > when I correct maths tests.
> > > > > > > > > >
> > > > > > > > > > do the primes you ask your students to
> > > > > > > > > > study fit in 64 bits?
> > > > > > > > > >
> > > > > > > > >
> > > > > > > > > Yes. But sometimes the primes I factor do
> > > > > > > > > not.
> > > > > > > >
> > > > > > > > Write a dc(1) script!
> > > > > >
> > > > > > man dc
> > > > > >
> > > > >
> > > > > I know, but why?
> > > >
> > > > Because it can already handle arbitrary precision
> > > > arithmetic.
> > > >
> > >
> > > I'm lost. Do you think factor(1) shall be implemented
> > > as a dc(1) script in sbase?
> >
> > No, as I already mentioned earlier, I don't want factor(1)
> > in sbase at this point in time. We can revisit this when
> > we have a dc implementation.
> >
> > This was just a suggestion at an alternative way of
> > implementing factor by just relying on existing POSIX
> > utilities and without having to implement your own bignum
> > code.
> >
>
> How are you going to implement dc without writing your
> own bignum code? Either, we use an existing bignum, write
> our own, or do not implement either of factor(1), dc(1),
> and bc(1). And if we are not going to implement them, then
> GNU factor will do.
>

dc is a general purpose language, factor isn't. If you want to focus
on something more meaningful for sbase, consider writing a small
bignum library that can be used with the future sbase dc. We can then
decide if factor should be implemented on top of dc or using this
bignum library.

In any case, dc is a needed utility for sbase, factor isn't so I can
provide no promises on factor inclusion.
Received on Fri Feb 26 2016 - 12:55:12 CET