New mpz prp functions...

[bodrato at mail.dm.unipi.it]

Ciao,

Il Mar, 9 Agosto 2011 3:08 am, David Cleaver ha scritto:
> Here are the prototypes for the new prp functions:

> int mpz_sprp(mpz_t n, mpz_t a) aka: Miller-Rabin pseudoprime

The Miller-Rabin test already is implemented in GMP in the file
mpz/millerrabin.c , and there is a function named mpz_probab_prime_p.
Did you consider how your function can be integrated with the existing
functions in the library?

> int mpz_strongbpsw_prp(mpz_t n)

I'd personally add at most one of them, possibly the strongest, if we can
integrate it with the current probab_prime_p.

> Again I'd like to ask, is there any formal procedure to ask for new
> functions to be included into GMP?

As far as I know, there is none. GMP is not a CAS, this means that it
should not contain lots of different functions, but only the basic ones,
so that anyone can build lots of new functions/programs using those
bricks.

We should ask to programmers using GMP if they need some new functions.
I'm sure that projects like ECM-GMP [ http://ecm.gforge.inria.fr/ ] or
PARI/GP [ http://pari.math.u-bordeaux.fr/ ] are implementing their own
(pseudo) primality testing algorithms. Do they need to have part of them
reimplemented in GMP? You may ask.

Regards,
Marco

-- 
http://bodrato.it/