I'm very sorry for the delay. :redface:

I guess I squashed the bugs and the stack overflow problem in version 4, but still need some Q&A before releasing it. I did not forget my promise. Stay tuned, it may come out anytime now... :showoff:

Luigi

A working release of Factor4 is now available.

The program has been tested on Windows XP and 98, though under Windows 98 there is a glitch related to the timing function.

It is delivered with its own libraries, with a zipfile of 640,088 bytes.

This time source code is included, so if you like to make Linux or MAC distros I'll happily include them on my webpage, giving credit to each one.

The zipfile is available [URL=http://www.moregimps.it/billion/factor4.zip]here[/URL]

Send me comments and complaints pls... :whistle:

Luigi

Benchmark is still raise stack overflow at me. Also, I couldn't resolve the idea of input.txt. When I use command-line: factor.exe < input.txt
I got:

**********************************************

Factor4 v. 4.00 - October 9th, 2005
(Mersenne numbers factorization written by Luigi Morelli)

Please enter the exponent to be factored: 2^
Now enter start bit depth : Finally enter end bit depth :
Sieving factors from 2.000 bits up to 1.000 bits...

Reinitializing sieve

Sieve time : 0.157 secs. sieve percentage : NaN

M195000023 no factor from 2.000 bits to 1.000 bits.

Total time : 0.296 secs.

*********************************************

Or this version doesn't do batch testings?

[QUOTE=HiddenWarrior]Benchmark is still raise stack overflow at me.
[/quote]

I'll dig into that.

[QUOTE=HiddenWarrior]Also, I couldn't resolve the idea of input.txt.
Or this version doesn't do batch testings?[/QUOTE]

This version doesn't do batch testing. For tests up to 59/60 bits and batch of 100 exponents it would cost a lot of work for a low increment of efficiency.
Finally, version 3_2 is less memory-hungry.

Luigi

[QUOTE=ET_]I'll dig into that.
[/QUOTE]

Ok, thanks!

If the Factor4 doesn't do batch-factoring then I guess we (100mdpp) will use factor 3_2 to the wide factoring.

Is there a way to make factor 3_2b to resume where I left off earlier? And if yes, how?

[QUOTE=HiddenWarrior]Ok, thanks![/QUOTE]

I recompiled factor4, and the actual version doens't segfaults on my machine.
I uploaded that version to let you try out.

Luigi

[QUOTE=OmbooHankvald]
Is there a way to make factor 3_2b to resume where I left off earlier? And if yes, how?[/QUOTE]

No, there isn't. You have to modify the input file, strip exponents that have been checked, and restart the application.

As each test up to 60 bits needs less than 30', I didn't think that a resume switch would be useful.

Luigi

New download page
 

There is a new download page at [URL=http://www.moregimps.it/billion/download1.php]Moregimps[/URL] .

Factor4.zip has source code in it.
If you like you can contribute your own distro for different OSes.

Luigi

Thanks! Benchmark now works good

Factor4 for Linux and AMD64!
 

Thanks to [B]Gribozavr[/B], the community can finally unleash the power of 64-bit Linux with Factor4! :bow: :bow:

You can download the binaries at [URL=http://www.moregimps.it/billion/download1.php]my site[/URL]

Waiting for other platform executables... :whistle: :help: :whistle:

Luigi