It seems I spent the last day working on a batchfile for naught. Brand new batch file, but somebody else is working on those same numbers and is just a bit ahead of me. Every number that I factor I find has been factored already.

I am a bit annoyed, but unfortunately that's the way it is right now and I understand that. Is there any way to lessen the chances of this happening again? Or should I just let others handle factoring?


Also, I've been working with yafu and I don't quite know what output file factorDB wants. I specified the -of flag in yafu to create a specific file for factors but what it outputs is not acceptable to factorDB. The text file yafu outputs to uses the form

[CODE](factored input number)/factor1/factor2
((1066001^17-1)/509008526182000)/1906325144800966919670787953338058650243696850751/3054621274906960872580871290784467972921[/CODE]

Help is appreciated. Apologies if this is the wrong place, but it seems you guys don't have the strictest of rules.

How do you fetch the composites which you factor?
How do you report the results, immediate after you factored a composite or do you report multiple at a ime later?

You can use [url=http://www.rechenkraft.net/wiki/index.php?title=Benutzer_Diskussion:Yoyo/factordb#yafu.pl]this[/url] Perl script together with the yafu binary to factor numbers from [url=http://factorization.ath.cx/stat_1.php]this[/url] list.

I retrieved the composites from the Downloads section of factorDB.

I was about to report the results I had for the entire day when I noticed that they had all basically been factored already. So, the latter, but this was my first time doing this work.


I've got the script running, thank you! It's very nice. There is still a chance of people doing work on the same number, but it seems to be slightly less likely.

I successfully directed yafu to gmp-ecm after some trial and error, but to what file do I direct it to for ggnfs? sieve.exe? If my ggnfs folder is C:\ggnfs, is that what I put in yafu.ini? It currently says:

ggnfs_dir=../ggnfs-bin/

Which is totally unhelpful considering I'm using Windows.

I have my ggnfs configured per Jeff Gilchrist's tutorial if that matters.

Hope I am being clear enough. I don't quite know all the terminology yet. I will respond again when I am better rested. :razz:

Find out where the gnfs-lasieve4I*.exe binaries are. (Probably C:\ggnfs\bin\) This will be your path in the YAFU.ini file. Don't forget to include the final backslash. Alternately, you can copy the six binaries into the YAFU directory.

Worked exactly as you said. I've got it all working splendidly now. Thank you both!

The database seems to be down.
edit: and up again

[QUOTE=henryzz;307703]Here is my modified script for ecm.
It has a bug somewhere and crashes/freezes early on sometimes. It only seems to effect early on so just restart it. I have run it for several hours successfully. In addtion if you get a prp*comp on the first number it will crash.[/QUOTE]

Tried running this script again on a new system and get errors at the beginning of the while loop. Can a perl buff tell me what is wrong easily?
I am using the latest version of perl that is in cygwin. This is 5.14.2.

Can you paste the screen output of the perl run?

[QUOTE=yoyo;331241]Can you paste the screen output of the perl run?[/QUOTE]

I have since managed to get it working. This version of perl needed (s escaping in ' ' blocks in this version. The previous version I used only needed it in " " blocks.
It was actually complaining on the next line not the line that had the bug just to be helpful.

Has someone been running this loads? I am finding so few factors I thought it wasn't working for a while. I have tried at several different digit levels.

[QUOTE=henryzz;331252]Has someone been running this loads? I am finding so few factors I thought it wasn't working for a while. I have tried at several different digit levels.[/QUOTE]

No, I have not been running it, but I would like the latest version.

I would also like [URL="http://mersenneforum.org/showpost.php?p=308082&postcount=3"]this post[/URL] addressed, or just a central repository for all the little varmints. :smile:

BTW, is [B]bchaffin[/B] still running his aliquot sequence helper script?

I noticed two issues when getting some aliqout seqs. These are lower priority as they are for non-open sequences.

[URL="http://factorization.ath.cx/index.php?id=239423790154880546"]239423790154880546[/URL] appears to have two factorizations that you get at random. One has: [URL="http://factorization.ath.cx/index.php?id=91054564999"]91054564999[/URL]
The other has the correct factorization for the above term.

Sequencs 580968 has an issue with index 148 according to aliqueit

[URL]http://factorization.ath.cx/index.php?id=316733334261546394[/URL] is correct, but the next sequence element appears to be wrong.

It should be 159012901563150950 given the factorization listed for the number above (3167....)
[URL="http://factorization.ath.cx/index.php?id=91054564999"][COLOR=#000000][/COLOR][/URL]