Re: GP as a programming language

[Ilya Zakharevich <ilya@Math.Berkeley.EDU>]

On Tue, Nov 19, 2002 at 04:30:34PM +0100, Bill Allombert wrote:
> > See the end of my previous message.  gp2c is in the same ballpark as
> > Math::Pari.  Math::Pari did not help GP/PARI.  I see no reason why
> > gp2c would.
> 
> gp2c allow to produce C code from GP code more easily. You do not need
> to care about the PARI syntax since gp2c handle this for you.

In my experience, tools to produce write-only code help in toy
situations only.  If we can't make C code which interfaces with PARI kernel
easier to read, nothing will help on the C front.

> If you have a working GP script you can use gp2c to translate it to C
> code and then add it to PARI. The file src/basemath/perm.c was done this way.
> So I believe it make extending PARI/GP easier.

My point was that with Math::Pari you can do extending as easy (or
easier) than with gp2c.

Ilya