about program NUMBERS

[flasholinos]

Hi quys ! im very new in this program and i want some help!!
when i install my program then start testing a M19922945 !! The program
gives numbers for testing or we must choose a number?? and how i can choose a number???? Is there a place which i can see which numbers is already tested and find a range to work??? THANKS

[S80780]

Hi, flasholinos!

If you run MPrime / Prime95 on a comp with internet - access, you can let the program ask a central server, the "PrimeNet" sever, for an exponent to test. You can let the sever choose which type of work your program gets or determine to get only special work types, namely Trial Factoring, Double Checking, First Time Tests or 10,000,000 digit Number Tests (You should only choose one of the last two, if you have a really fast comp or a lot of patience!). In general, it's the best to let the server choose which type of work your comp gets.
If your comp doesn't have internet access, you can also use the Manual Test Form to get work and report results.

[trif]

Very odd, that exponent isn't even prime. Primenet couldn't have given it out. I'm not even sure Prime95 would LL test it since it isn't a prime exponent.

[S80780]

I just tried it out. Program output was:
"Error: Work-to-do file contained composite exponent: 19922945"
I expect the primality test is the same as in decomp.c:
[code:1]
/* Return TRUE if arg is a prime */

int isprime (unsigned long p)
{
unsigned long f, fmax;
if ((p & 1) == 0) return (p == 2);
if (p % 3 == 0) return (p == 3);
fmax = sqrt (p);
for (f = 5; f <= fmax; f += 6) {
if (p % f == 0) return (0);
if (p % (f+2) == 0) return (0);
}
return (1);
}
[/code:1]
Benjamin

[garo]

I think there must be a typo in the original statement.

[eepiccolo]

Originally Posted by garo

I think there must be a typo in the original statement.

Maybe, but who knows.

Anyway, the questions that flasholinos asks are answered in the help file. And there is much more info other than that in the help file also.

[wpolly]

It Maybe The Self-Test of the program.
AFAIK, The Self-Test of 1024K FFT Always Begins with M19922945.

[Khemikal796]

But why M19922945, because its an easy devide by 5? Or is a simple factor like this picked for a reason?

[trif]

Originally Posted by wpolly

It Maybe The Self-Test of the program.
AFAIK, The Self-Test of 1024K FFT Always Begins with M19922945.

Aha! The mystery is solved! :D

[ewmayer]

Originally Posted by Khemikal796

But why M19922945, because its an easy devide by 5? Or is a simple factor like this picked for a reason?

For self-testing purposes it doesn't matter whether the exponent is prime or not - the Mersenne-mod DWT algorithm works the same way in either case. Now if you saw a composite exponent slated for LL testing in your worktodo file, now that would be worrisome.