Prime Pages How to's

[PawnProver44]

Originally Posted by a1call

Why not register and login before submitting primes?
That way it will be registered under your name rather than others and kept track of here:
http://factordb.com/certtop.php

I can easily find these primes, so it is no big deal. I found 5 today.

[PawnProver44]

Originally Posted by paulunderwood

There is nothing to stop you doing this. It will take much longer to do each individual test of numbers where on average the primes are more spread apart, plus you will be using generic modular arithmetic of the underlying library which slows things down further. However, you might get lucky by striking gold on your first attempt

To prove a number prime, should it be factored to 33.33% or 12.5%. I read many sources saying each percentage is fine.

[PawnProver44]

I was originally going to start a new thread, but I think I created enough for a week so I just decided to leave it in this one:

I am trying to prove a number prime with a primality certificate, however I have two problems. One, WolframAlpha wolframalpha.com (link) can prove numbers prime less than 10^2999, with certificates, and there is no way to upload any links to factordb.com. Second is that I can use Primo, but I can't run it on Windows 8. Is there any solution to this? I found a PRP (well actually prime) with WolframAlpha, and factordb says it is a PRP, while WolframAlpha says it is prime (Wait 10-20 seconds on WolframAlpha and you will see):

http://www.wolframalpha.com/input/?i...5E860-31+prime

http://factordb.com/index.php?query=59*46%5E860-31

Also, Prime Pages would not accept this prime, why is that?

[Uncwilly]

INB4MM

[paulunderwood]

Originally Posted by PawnProver44

To prove a number prime, should it be factored to 33.33% or 12.5%. I read many sources saying each percentage is fine.

33.33...% of N-1 is easy. Slightly bigger than 12.5% of N^2-1 requires more time to prove but is doable.

[paulunderwood]

Originally Posted by PawnProver44

Also, Prime Pages would not accept this prime, why is that?

We at the Prime Pages attempt to keep a list of the 5000 largest known primes plus a few each of certain selected archivable forms. To make the top 5000 today a prime must have 388341 digits or meet the size requirements for it's archivable form. (Query time: 0.001815 seconds.)

Originally Posted by PawnProver44

I am trying to prove a number prime...

What number? Maybe you just need to learn about PFGW helper files.

[a1call]

To run Primo on windows:
* Download and install free virtual box
* Download and install 64 bit ubuntu
* Download and install Primo

[danaj]

Originally Posted by PawnProver44

WolframAlpha wolframalpha.com (link) can prove numbers prime less than 10^2999, with certificates, and there is no way to upload any links to factordb.com.

(1) are you sure your link does a proof? How can you tell?

(2) PrimeQ doesn't do a proof.

(3) wolframalpha.com does not recognize any of the functions from the Primality Proving package. What is the command you're using to make it call PrimeQCertificate[n]?

(4) factordb.com accepts only Primo certificates. Accepting more certificate types would be both good and bad. It would be nice to allow other tools, but it also means people looking up numbers would find certificates they couldn't immediately verify or parse. By restricting to Primo, there is only one format (which is far better documented than Wolfram's) and three verifiers.

Wolfram's documentation for certificates is pretty crappy IMO. You have to wander through a number of web pages and are still left without clear instructions. They actually do a good job of showing how the certs work for compositeness, but then just wave their hands around the rest. The Pratt portion can be muddled through if you use Mathworld and some reverse engineering. The ECPP certificate is not shown, so good luck actually verifying it. The Mathworld page gives hints as to what to expect, so I'm pretty sure it could be reverse engineered, but that's rubbish. "Here's my certificate of primality: 'squeamish ossifrage' and here's my closed-source function that verifies it." Why would anyone want to use this?

[schickel]

Originally Posted by PawnProver44

Second is that I can use Primo, but I can't run it on Windows 8. Is there any solution to this?

If you look over in the "FactorDB" sub-forum, there is some discussion about Primo. If you have the Windows version, it *might* run on Win8, but no guarantees. Best solution is either a Linux box or run Primo on a VirtualBox VM.

http://factordb.com/index.php?query=59*46%5E860-31

You've got a few options: someone might prove it on the way to proving all PRPs below 3,000 digits, prove it yourself if you get Primo running, or talk someone into moving it up a little higher on their priority list to prove it. (Right now, I would have to pause my c158 that is sieving on my bigger machine to run it or let it stew on the little one with the Windows Primo...)

[PawnProver44]

Originally Posted by paulunderwood

What number? Maybe you just need to learn about PFGW helper files.

Is 59*46^860-31 (proven) prime? I have two sources one saying it is a PRP (factordb.com) and one saying it is a prime (wolframalpha.com). I think it is prime, but two sources are saying opposite statuses.

[PawnProver44]

[QUOTE=We at the Prime Pages attempt to keep a list of the 5000 largest known primes plus a few each of certain selected archivable forms. To make the top 5000 today a prime must have 388341 digits or meet the size requirements for it's archivable form. (Query time: 0.001815 seconds.) [/QUOTE]

I read the entire site, but do not explain if you can submit prime neither of Achievable forms, or greater than 388341 digits. They quickly deleted a few small primes around 1500 digits I tried to submit. I just wanted them to be in the database, not necessarily the list. That's all.