Aliqueit.exe discussion - Page 7

Joshua2 · 2009-03-30, 18:34

does aliquot use msieve for the final calculations, or for the initial polynomial selection? I have read that msieve is better at both those tasks.

mklasson · 2009-03-30, 18:42

Quote:

Originally Posted by Joshua2

does aliquot use msieve for the final calculations, or for the initial polynomial selection? I have read that msieve is better at both those tasks.

By default it's set up to use msieve for gnfs postprocessing, yeah. That's all done outside of aliqueit though. Just make sure aliqueit.ini->gnfs_cmd uses factMsieve.pl and it'll be alright.

Alas, no, it uses the polyfinder in ggnfs for now. I've considered letting msieve find a poly, but haven't implemented it yet.

Andi47 · 2009-03-30, 19:27

Quote:

Originally Posted by mklasson

Alas, no, it uses the polyfinder in ggnfs for now. I've considered letting msieve find a poly, but haven't implemented it yet.

Is it possible to start an ubasic program from the "DOS"-command line? In this case, the line calling facmsieve.pl could be (optionally!) replaced by a line calling my Ubasic script for GNFS*), which uses msieve for poly selection and postprocessing.

*) the script needs a worktodo.txt file to start. It needs renamed gnfs-lasieve4I1?e files too, see the documentation of gnfs.ub. I think the script still contains the line which copies msieve.log to msieve.old when it's finished, so the factors can be parsed out of the msieve.old file.

10metreh · 2009-03-30, 19:29

Quote:

Originally Posted by Andi47

Is it possible to start an ubasic program from the "DOS"-command line? In this case, the line calling facmsieve.pl could be (optionally!) replaced by a line calling my Ubasic script for GNFS*, which uses msieve for poly selection and postprocessing.

(the script needs a worktodo.txt file to start. It needs renamed gnfs-lasieve4I1?e files too, see the documentation of gnfs.ub. I think the script still contains the line which copies msieve.log to msieve.old when it's finished, so the factors can be parsed out of the msieve.old file.)

Personally, I think an option in the config file would be best for this, because gnfs.ub is susceptible to various odd ubasic bugs.

Joshua2 · 2009-03-30, 19:32

I believe it doesn't work with vista x64 as well I heard, which it wouldn't if its 16bit.

Andi47 · 2009-03-30, 19:38

Quote:

Originally Posted by 10metreh

Personally, I think an option in the config file would be best for this, because gnfs.ub is susceptible to various odd ubasic bugs.

That's what I was thinking of when I was writing the word "optionally!".

P.S.: theoretically it should be possible to use gnfs.ub without a change in aliqueit.exe - for this a change in gnfs.ub would be necessary to output the factors in the same way as factmsieve.pl does, and to change the line in the .ini file to call gnfs.ub instead of factmsieve.pl.

mklasson · 2009-03-30, 19:44

Quote:

Originally Posted by Andi47

Is it possible to start an ubasic program from the "DOS"-command line? In this case, the line calling facmsieve.pl could be (optionally!) replaced by a line calling my Ubasic script for GNFS*), which uses msieve for poly selection and postprocessing.

Beats me. But yeah, all aliqueit needs from the program run with gnfs_cmd is a ggnfs.log file afterwards with the factors presented in msieve or factLat.pl fashion. As input to gnfs_cmd aliqueit writes a test.n file containing "n: <composite>".

edit: oh, and it passes "test" as argument to gnfs_cmd.

Andi47 · 2009-03-30, 19:57

Quote:

Originally Posted by mklasson

Beats me. But yeah, all aliqueit needs from the program run with gnfs_cmd is a ggnfs.log file afterwards with the factors presented in msieve or factLat.pl fashion. As input to gnfs_cmd aliqueit writes a test.n file containing "n: <composite>".

So I need to tell gnfs.ub that test.n is a valid input file. I will try to get it to run during my easter holiday.

Joshua2 · 2009-03-30, 20:05

I could probably write a c# exe to run msieve and convert and make a proper .poly so gnfs can just run, would that help?
If aliquot used 1.40 instead of 1.38 would that speed things up much? Or maybe thats yafu's fault, but does it make much difference?

Andi47 · 2009-03-30, 20:19

Quote:

Originally Posted by Joshua2

If aliquot used 1.40 instead of 1.38 would that speed things up much? Or maybe thats yafu's fault, but does it make much difference?

Using msieve 1.40 instead of anything older than msieve 1.39 would speed up GNFS *very much*, as msieve 1.39 and newer finds MUCH better polynomials than msieve 1.38 and older.

Joshua2 · 2009-03-30, 20:20

I think I was referring to yafu using msieve 1.38 in post processing. Is that changed?

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Resuming aliqueit	johnadam74	Aliquot Sequences	4	2016-03-28 12:32
Apparent aliqueit issue with specifying factors	pakaran	Aliquot Sequences	2	2015-09-12 23:10
Using Several Instances of Aliqueit for a large gnfs job	EdH	Aliquot Sequences	6	2011-12-13 18:58
Setting up aliqueit	science_man_88	Aliquot Sequences	185	2011-11-08 12:18
Tried out aliqueit.exe: ggnfs failing	Greebley	Aliquot Sequences	35	2010-02-13 15:23

2009-03-30, 18:34	#67
Joshua2 Sep 2004 215₁₆ Posts	does aliquot use msieve for the final calculations, or for the initial polynomial selection? I have read that msieve is better at both those tasks.

2009-03-30, 19:32	#71
Joshua2 Sep 2004 215₁₆ Posts	I believe it doesn't work with vista x64 as well I heard, which it wouldn't if its 16bit.

2009-03-30, 20:05	#75
Joshua2 Sep 2004 13·41 Posts	I could probably write a c# exe to run msieve and convert and make a proper .poly so gnfs can just run, would that help? If aliquot used 1.40 instead of 1.38 would that speed things up much? Or maybe thats yafu's fault, but does it make much difference?

2009-03-30, 20:20	#77
Joshua2 Sep 2004 13·41 Posts	I think I was referring to yafu using msieve 1.38 in post processing. Is that changed?