// I am seriously impressed by how PFGW handles this number, which // is the sum of 3 cubes, each with more than 6 million digits. // I recommend this as a...
This file craps out the Win32 build :( It is GMP dieing (am am 99% sure). I will look at it a little, and I have placed the file (with your comments) into...
I have spent a few days overhauling the PFIO stuff (and also some in the PFOO). What I have done is to "decouple" the PFIO from PFOO (and PFGWLIB, and most...
I think it is high time that we try to get a solid release (1.2) out the door. I have made some significant changes recently, that are going to require a...
I was assuming you would break it (as usual). However, it is best if that is done in a coordinated manner, so that we "know" what things you tried to break,...
... PS: Maybe only that funny grey-screen version that some folk use? I got it through pfgw.exe (vintage June 2004) under Windoze95 on a old 266MHz Pentium2,...
Silly me. I have a libgmp-3.dll in the path which uses alloca. Once I changed that (cleared my path), pfgw worked just fine. Takes about 10 seconds on my...
Thanks for getting to the bottom of that, Jim. I had been wondering why pfgw.exe let Rosinante handle this input, while Winpfgw.exe did not. The reason was the...
... No. The correct libgmp-3.dll can speed up processing (at times quite a bit). However, these DLL's have been compiled using alloca (since it speeds up gmp...
Thanks Jim, thanks Jim, thanks Jim, thanks. What I really wanted was to perform 30-million-digit Lucasian evaluations (no PrP-ing, of course) on a 128 MB...
Hello, This email message is a notification to let you know that a file has been uploaded to the Files area of the openpfgw group. File :...
openpfgw@yahoogroups....
Jul 5, 2004 2:50 pm
1541
Hello, This email message is a notification to let you know that a file has been uploaded to the Files area of the openpfgw group. File :...
openpfgw@yahoogroups....
Jul 5, 2004 2:50 pm
1542
I made ... my working version on a couple of Windoze machines. One is a 1GHz Athlon that is temporarily idle, so I could do some serious testing, if someone ...
Two earlier reported bugs are still in PFGW 20040704. ... The prime generator produces (all?) composites without a factor below 2^15. This means the first...
... below 2^15. This ... rare cases with ... more (by 31 ... Fixed, although you might not like the fix. Simply put, p(x) has been changed to only generate...
... composite with ... v23 w/P4] ... Fixed, Now within gw_prp.cpp, we first check to see if N%iBase==0 (only doing so if N is 1 gmp limb or less) Broken PFGW:...
... I was going to say that such a small answer can't be the 100010000-th partition number, as that's the usage I'm most used to for p() (for example the...
This contains the fixes mentioned today 1. Fermat test does not allow base divisible by the number. 2. the primegen p(x) does not generate primes over (2^15)^2...
Hi! The last week I tried to implement a prime generator in order to do a trial division on numbers of the form 2*n!-1. My generator is nothing special: It's...
... OpenPFGW is fast to prp large numbers but the trial division uses a simple algorithm and can be improved a lot. My GMP based TreeSieve 1.1 is here: ...
... Not only does it use a simple algorithm, but the trial factoring function is actually PART of the expression parser (PFFunctionSymbol) Most of the speed...
... This is due to the asm code in PFGW. The code is a true O(N^2) "basecase" function. If using GMP division, there is a basecase, a Karatsuba, ToomCook,...
Some things to keep in mind about PFGW's factoring code. A. N is divided by each candidate factor prime, one at a time. B. The result is checked to see if N...
Thanks, Phil! ... Next question: why are some powers a lot worse than others? If I didn't boob, there are 134 base-64 pseudoprimes less than 10^4, but only 71...
... The method is not usual the Euler-GCD-tree, all residues are kept, so you're alright. It should be a trivial change. I can't say I find Jens' code...
Having problems with message search?
Fill out this form to ensure your group is one of the first to be migrated to the new message search system.
