M4219 completely factored?

[TimSorbet]

Originally Posted by WVU Mersenneer

May I ask how you use Prime95 for PRP?

Here's the line to put in worktodo.txt:

/*    PRP=k,b,n,c[,how_far_factored,tests_saved][,known_factors]    */
/* A tests_saved value of 0.0 will bypass any P-1 factoring        */

If I'm not mistaken, the how_far_factored and tests_saved are so it can calculate how much, if any TF/P-1 to do prior to running it, just like with a Test= line. known_factors must be inside double quotes, (so it's easily distinguished from the other parts of the line) and can include multiple factors separated by commas. e.g. PRP=...,"1234567" or PRP=...,"1234567,891234567". In case you're not familiar with it, the "k,b,n,c" mean the number you're testing is k*b^n+c (divided by any known factors). For Mersenne numbers/cofactors, these are "1,2,p,-1", (with the exponent in p, of course) to make it the same as 2^p-1

[MatWur-S530113]

Originally Posted by WVU Mersenneer

Sounds like we're in the same boat trying to get GMP-ECM to work on Windows! The ease and speed with which Prime95 runs on Windows is further testament to Dr Woltman and his band of experts behind the scenes.

Yes, my problems installing GMP/GMP-ECM are ver hindering. At least Prime95 works well, thus I can work a little bit on the Mersenne's- and the Fermat's-numbers.

Originally Posted by WVU Mersenneer

May I ask how you use Prime95 for PRP? I tried earlier this week by simply removing the largest found factor for an exponent and running curves on the exponent at a level at least 2 above the size of the factor, for example if the factor was 40 digits, I ran curves with B1=44e6, but most of the factors were much smaller. However, this barely worked at all as ECM is not a "sure thing" by any means and has often amazed me at how many curves need to be run in order to find a factor...as well as how few sometimes.

nono^^
as Mini-Geek already said:
1) stop prime95
2) add a line like
PRP=1,2,1108369,-1,0,0,"24384119"
to worktodo.txt
3) restart prime95

this will run a prp test on the number 1*2^1108369-1 (= M1108369) using the already known factor(s) 24384119. In the result file you will get an output line like:
UID: Wurtinger/i5-750, M1108369/24384119 is not prime. RES64: 8D485CB5ABCC37DA. Wd4: 9F2E0565,00000000

(step 2) you can add a line like
PRP=1,2,1108369,-1,"24384119"
to worktodo.txt to do the prp-test, prime95 automatically expands the input by ',0,0'

Originally Posted by WVU Mersenneer

Again, thank you for your help and I am eager to know what you find.

You are welcome and no PRP found so far^^

greetings
Matthias

[WVU Mersenneer]

Matthias and Mini-Geek, I am indebted to you both. I sincerely thank you for taking the time to provide me with this information as well as for your candor and support throughout this thread.

[Brian Gladman]

Can those who are having trouble with GMP-ECM on Windows please indicate what problems they are having?

The SVN repository has all the files needed to build GMP-ECM with Visual Studio 2010 (using MPIR) for both 32 and 64 bit Windows systems. These builds work for me and I believe that Jeff Gilchrist has working copies of the binaries here:

http://gilchrist.ca/jeff/factoring/index.html

Brian

[WVU Mersenneer]

Originally Posted by Brian Gladman

Can those who are having trouble with GMP-ECM on Windows please indicate what problems they are having?

The SVN repository has all the files needed to build GMP-ECM with Visual Studio 2010 (using MPIR) for both 32 and 64 bit Windows systems. These builds work for me and I believe that Jeff Gilchrist has working copies of the binaries here:

http://gilchrist.ca/jeff/factoring/index.html

Brian

Brian, please know I meant in no way to denegrate GMP-ECM or anyone associated with it, and if I came across that way I apologize. Even though my career is in computers and my passion is in numbers, I am grostesquely awful at understanding and implementing both.

I tried a couple of months ago to download GMP-ECM based upon a thread I can no-longer recall. I was able to find the files, download and extract them, and even go through the steps of installing it on my PC, however, it took nearly 10 hours for everything to appear to be finished, and I'm not sure it actually was. From there I have been unable to get the window open again in order to attempt to do anything with it.

I'll state outright that I probably did everything wrong from the start and was doomed to fail all along. I will follow the link you provided, thank you for it btw, and let you know as specifically as I can what difficulties I encounter.

[Brian Gladman]

Rest assured that I didn't your comments as a criticism so please don't worry about that. I am interested in understanding your problems in case the is an issue that I can do something about.

There is no way it should take 10 hours to install GMP-ECM (unless you are doing it on a Ferranti Pegasus or a English Electric KDF9) so I would suggest you delete what you have and start again. It isn't easy from source but it should work with the binaries that Jeff provides (they may be a bit out of date).

If you want some installation guidance on Windows x64 please PM me with an email address and I will try to answer any questions you have.

Brian

[MatWur-S530113]

Originally Posted by Brian Gladman

Can those who are having trouble with GMP-ECM on Windows please indicate what problems they are having?

The SVN repository has all the files needed to build GMP-ECM with Visual Studio 2010 (using MPIR) for both 32 and 64 bit Windows systems. These builds work for me and I believe that Jeff Gilchrist has working copies of the binaries here:

http://gilchrist.ca/jeff/factoring/index.html

Brian

I have trouble installing GMP-ECM because I have trouble to install GMP.
I don't have Visual Studio 2010, only the free version of Visual C++ I have downloaded and installed (with the hope, that it is enough to compile the libraries).
Some years ago there was a thread with a link to a site with a step-by-step instruction to install GMP and GMP-ECM with MinGW. With this help I was able to get a working GMP/GMP-ECM installation on my old Computer. Now this site is not longer available and it seems impossible to me to compile only the GMP-library... I don't know where to start, if I try to open a project-file with Visual C++ then I don't find such a file in the (unzipped) GMP-folder. And if I will find such a file, where i have to choose (or type in) the '$make' command? I'm not a C++-programmer, that's why I have the trouble. And I don't find anywhere a site with instructions for non C++-programmer how to install GMP and which programs I need to do it (Visual C++? MinGW? CygWIN? MPIR?). That's (strong shortend) are my problems to compile GMP and, as a 'follow up', to compile GMP-ECM.

greetings
Matthias

[axn]

Originally Posted by MatWur-S530113

I have trouble installing GMP-ECM because I have trouble to install GMP.
I don't have Visual Studio 2010, only the free version of Visual C++ I have downloaded and installed (with the hope, that it is enough to compile the libraries).
Some years ago there was a thread with a link to a site with a step-by-step instruction to install GMP and GMP-ECM with MinGW. With this help I was able to get a working GMP/GMP-ECM installation on my old Computer. Now this site is not longer available and it seems impossible to me to compile only the GMP-library... I don't know where to start, if I try to open a project-file with Visual C++ then I don't find such a file in the (unzipped) GMP-folder. And if I will find such a file, where i have to choose (or type in) the '$make' command? I'm not a C++-programmer, that's why I have the trouble. And I don't find anywhere a site with instructions for non C++-programmer how to install GMP and which programs I need to do it (Visual C++? MinGW? CygWIN? MPIR?). That's (strong shortend) are my problems to compile GMP and, as a 'follow up', to compile GMP-ECM.

greetings
Matthias

No offense, but why are you so hell bent on compiling from source, when you don't have the tools to do so? Why can't you just download the pre-built binaries and run them?

[Brian Gladman]

Originally Posted by MatWur-S530113

I have trouble installing GMP-ECM because I have trouble to install GMP.
I don't have Visual Studio 2010, only the free version of Visual C++ I have downloaded and installed (with the hope, that it is enough to compile the libraries).
Some years ago there was a thread with a link to a site with a step-by-step instruction to install GMP and GMP-ECM with MinGW. With this help I was able to get a working GMP/GMP-ECM installation on my old Computer. Now this site is not longer available and it seems impossible to me to compile only the GMP-library... I don't know where to start, if I try to open a project-file with Visual C++ then I don't find such a file in the (unzipped) GMP-folder. And if I will find such a file, where i have to choose (or type in) the '$make' command? I'm not a C++-programmer, that's why I have the trouble. And I don't find anywhere a site with instructions for non C++-programmer how to install GMP and which programs I need to do it (Visual C++? MinGW? CygWIN? MPIR?). That's (strong shortend) are my problems to compile GMP and, as a 'follow up', to compile GMP-ECM.

You can build GMP-ECM with Visual C++ Express 2010 but if you want the 64-bit version you also have to install the Windows 7.1 SDK (another free Microsoft download). You also need to install Python (2.6 or later).

You should start by downloading and unzipping the latest MPIR version (from www.mpir.org). Once unzipped, you will find a readme.txt that describes how to build MPIR using VC++ Express.

Once you have built MPIR you can then build GMP-ECM (the MPIR and GMP-ECM directories need to be side by side). GMP-ECM is not currently set up to build with Express but if you get this far, I can describe a few simple steps that are needed to convert the build files so that they work with Express.

I wouldn't call this an easy process so you might want to start by downloading a prebuilt binary from Jeff Gilchrist's site.

Brian

[ATH]

I ran a strong BPSW primality test on M86137/7747937967916174363624460881 and it came back as composite, which should be definite. All probably prime test I heard of can only fail by showing a composite number as prime, not a prime as composite.

Then I tried a Prime95 PRP test which I should have done first, but I thought someone else in this thread did it:

M86137/7747937967916174363624460881 is not prime. RES64: B338C6AC7B9C465B. We4: 81405965,00000000

So I guess it isn't prime after all.

[Prime95]

Originally Posted by ATH

M86137/7747937967916174363624460881 is not prime.

So I guess it isn't prime after all.

You need to run the PRP test dividing M86137 by BOTH known factors