Are there any plans to implement save files for PFGW, to allow resuming in the middle of testing a particular number? I know that the usual use of PFGW is for...
1859
Michael Bell
mdb36@...
Apr 23, 2001 4:08 pm
Hi, I reckon that it should take about 2 and a quarter P3-800 CPU years to do 6000 PRP tests. Currently I don't think PFGW can save intermediates during a...
1860
Arlin Anderson
arlin@...
Apr 23, 2001 5:07 pm
... From: Michael Bell [mailto:mdb36@...] Sent: Monday, April 23, 2001 4:59 AM ... You should have about P=0.9 of getting at least 1. Yours, Arlin...
1861
Arlin Anderson
arlin@...
Apr 23, 2001 5:27 pm
... From: Michael Bell [mailto:mdb36@...] Sent: Monday, April 23, 2001 11:09 AM ... After each, say 100th or thousandth iteration (out of more than...
1862
jfoug@...
Apr 23, 2001 6:06 pm
... Sounds about right. That is why the statement "20 PC's for 2 weeks" gets us into the "possible"; range. That is around 20% or so of expected. Running this...
1863
Phil Carmody
fatphil@...
Apr 24, 2001 1:30 pm
On Tue, 24 April 2001, primeform@yahoogroups.com wrote: From: "Michael Bell" <mdb36@...> ... The 0.56/ln(p) is an approximation for big p. It's _simpler_...
1864
paulunderwood@...
Apr 25, 2001 3:25 pm
Hi, how much longer do you think sieving will take? I look forward to the output. Perhaps you could break it up into digestible chunks for us, zip it up and...
1865
jfoug@...
Apr 25, 2001 5:31 pm
... the ... Yes, please cut out 5000 or so candidates, and sieve these to a high level first, and then pickup sieving on the remaining group. That would allow...
1866
jfoug@...
Apr 25, 2001 5:41 pm
The database section should work fine. If we get some data to start with, I can get the database records loaded (I hope I can import correctly) How about...
1867
Arlin Anderson
arlin@...
Apr 25, 2001 5:42 pm
... From: jfoug@... [mailto:jfoug@...] Sent: Wednesday, April 25, 2001 12:31 PM ... It is more efficient to sieve a large batch, than two smaller...
1868
jfoug@...
Apr 25, 2001 5:58 pm
Is it so with this sieve? I know with fixed k sieving (I believe the sieve for this is similar), you do lose a little by reducing the range but you can push...
1869
Chris Nash
chris_nash@...
Apr 25, 2001 10:49 pm
Hi folks ... There are currently two reserved p prover codes, p0 and p1, that have the "entity" PrimeForm as their only member. p0 has never been used and was...
1870
d.broadhurst@...
Apr 25, 2001 11:18 pm
... I'm rather proud to be a member of that team. Each time I see an F2 "credit" divided by 4, I ask myself: am I really worth 25%? David...
1871
Chris Nash
chris_nash@...
Apr 25, 2001 11:42 pm
Hi folks, I noticed Jim has already populated a 100,000 NRD search database in the group area. We've never used this before, since it really is awkward to use,...
1872
d.broadhurst@...
Apr 25, 2001 11:48 pm
A thought: before anyone modifies a table s/he could take a copy for safe-keeping?...
1873
Michael Bell
mdb36@...
Apr 26, 2001 12:57 am
Hi, The sieving program only works up to 2^31 anyway (at the moment). Its currently at 1375329203, and should reach 2^31 before then end of the week. I will...
1874
Nathan Russell
nrussell@...
Apr 26, 2001 3:43 am
... Agreed! I've completed four runs (since I was running Linux some of the time, and run strictly GIMPS there since I can't switch projects quite as quickly)...
1875
Chris Nash
chris_nash@...
Apr 26, 2001 4:32 am
Hi folks... FYI.... The ratios for pfgw should be about the same - remember this is per- iteration for Mersenne (multiply by 3 for pfgw), and the 512k FFT ...
1876
Phil Carmody
fatphil@...
Apr 26, 2001 9:53 am
... You see David, it's 25% because of the APR-CL. With just BLS it would be 33 1/3% ! Phil (yes I'm just being silly) Mathematics should not have to involve...
1877
Christ van Willegen
christw@...
Apr 27, 2001 10:29 pm
Hi guys/gals, I seem to have a bit of a problem with my .INI file. My computer crashed,and PFGW was unable to write the .INI file to disc. It measured 0...
1878
jfoug@...
Apr 28, 2001 12:30 am
I have a few questions before I go digging: What OS? What version of PFGW? What type of file (Newpgen, ABC, ABC2, "arbitrary")? I thought the restarting was...
1879
Christ van Willegen
christw@...
Apr 28, 2001 8:03 am
Sorry Jim, my bad, I should have given you this immediately: Win2000 PFGW Version 20010415.Win_Dev ABC2 Here's the content of the .INI file: [PFGW] MODE=PRP ...
1881
d.broadhurst@...
Apr 28, 2001 6:44 pm
I have was relieved that Carl Pomerance did not object to 3*F1+F2>1. My source is the theorems quoted in Tony's VFYPR.C. PFGW now does what VFYPR does! But it...
1882
d.broadhurst@...
Apr 28, 2001 7:09 pm
PS: http://groups.yahoo.com/group/primeform/files/N^2-1/G148.txt is a log of VFYPR doing a pure BLS proof with 3*F1+F2 > 1 > 3*F1. In essence, Chris N has...
1883
Michael Bell
mdb36@...
Apr 28, 2001 7:32 pm
Hi, Files are uploaded! Find 34k.txt - 46k.txt in the 100000 NRD directory in the files area. Sign them out using the database so noone does the same work...
1885
d.broadhurst@...
Apr 28, 2001 10:01 pm
Marcel Martin ... Interesting. Can you give the theorem or a link to it? The problem does not arise with APR-CL, since S is constructed from small primes that...
1887
Chris Nash
chris_nash@...
Apr 29, 2001 3:35 am
Hi there folks ... the ... Currently, PFGW does not use the factor bound B in its combined test. This is beginning to annoy me, because although 3*F1+F2>1 is...
1888
d.broadhurst@...
Apr 29, 2001 11:49 am
... Thanks for explaining the guts of the PFGW -tc test, Chris. The neat thing, as you say so clearly, is that ... What is apparent from our musings is that...
1889
d.broadhurst@...
Apr 29, 2001 12:09 pm
Offline I get lots of comments applauding the ingenuity of recent Lucas/Fib proofs. I want to say publicly that Bouk is the powerhouse behind these. In the...
1890
d.broadhurst@...
Apr 29, 2001 1:08 pm
... Well I don't lay claim to be a standard human, but I did succeed in compressing the whole of APRCL into 23 lines of Pari source. In ...
