Sieving drive for k=2000-3400 n=50K-1M

[Lennart]

15T-16T done and emailed
Lennart

[Lennart]

Reserving 23200G-24T Lennart ETA June 12

[Lennart]

16050G-19T complete and mailed to Gary

/Lennart

[Flatlander]

7.5T-7.9T is complete. (I'll email Gary.)

[gd_barnes]

Lennart or Max,

Can you send me the current sieve file with the n=50K-400K range removed in an Email or provide me with a link to it?

I'm going to use the remaining n=400K-1M range to compute optimum sieve depth.

Lennart, as fast as you're going through the ranges, if we can make it to P=30T sometime between the 15th and 20th, I may suggest doing that; assuming that the optimum is P>25T using the above range. (I think it will be.) If we do that, I could likely do a final P=500G range that would complete in 3-4 days.

Everyone,

I know it seems like we keep shooting for a moving target here but in this case, it clearly makes sense to be flexible. The target was more time-based than it was sieve-depth based (because I figured we'd be in the ballpark by the 20th) to get some lower top-5000 tests to begin searching on the servers but if the math says to go deeper, we'll do that since it's within the original time constraint.

It's amazing the large efficiency gain by sieving this much larger k and n-range all at once. None of our previous drives had n<=600K sieved above P=20T. Back when the project first started for k=400-1001, optimum for n=260K-600K was P=5T on 32-bit machines. For k=1003-2000, we sieved n=50K-500K and n=500K-1M completely separately and were intent on "breaking off" n=500K-600K and so only used it's optimum search depth, which was in the P=~20T range. I had no clue that we would blow throught that range so fast or I would have had them sieved together. I determined then that from here on out, all k-ranges would get sieved in one big huge chunk for n=20K (or whatever) to n=1M.

Now with this drive, even if you remove the n=50K-400K range, the much larger 1400k (vs. 1000k or 600k) k-range and n=600K (vs. 400K or 500K) n-range makes for an optimum depth even higher; likely P=25T-30T for LLRing up to n=600K. It may be P>60T for n=600K-1M. (I stopped k=1003-2000 at P=45T since we are only LLRing k=1400-2000.)

It's amazing how far the software and machines have come in just 18 months!


Gary

[Lennart]

20T-21T done and mailed to Gary

Reserving 24T-25T ETA June 13

/Lennart

[gd_barnes]

Originally Posted by Svenie25

12400G-12650G done. Taking 23000G-23200G, ETA june 14th.

Sven,

I am getting an error when I attempt to pull up your file using WinRAR or WinZIP. It's an unusual error that I have never encountered before. Here it is:

!   C:\Users\test\AppData\Local\Microsoft\Windows\Temporary Internet Files\Content.IE5\CA5ETZDF\factors%2012400-12650[1].zip: Unknown method in factors 12400-12650.txt

I wonder if there is some unusual problem with a carriage control character that is making the file unreadable.

Can anyone attempt to pull up Sven's file? If someone can get it pulled up, please save it off and Email it to me.

Sven, if neither you nor anyone else can pull the file up from your posting here, I'm going to have to ask you to Email me another one unzipped. There must be something wrong with your zipping process. I tried this 4 times and got the same error each time.

Edit: I googled this error and it says I may not have the latest version of WinRAR, which I did not. I upgraded to the latest version but am still getting the same error.

Edit2: I just now tried this on one of my Linux machines. It says: "An error occurred while extracting files." In another box, it says "unsupported compression method 98".

Clearly a compression method has been chosen that even some of the newest common compression software is unfamiliar with. After 20-30 mins. of messing with this, if anyone can enlighten me or just send the file to me, that would be great. Better yet, the file is small enough that it doesn't need to be zipped if sent in an Email. Just attaching the full file to an Email would work great.


Thanks,
Gary

[gd_barnes]

I just now got the file from Karsten. Karsten, what compression software are you using? I'm still baffled.

[Svenie25]

Originally Posted by gd_barnes

I just now got the file from Karsten. Karsten, what compression software are you using? I'm still baffled.

I used WinZip 12.1 for this file, I think. Very strange that you become an error. For the next files I will use again the normal method and email you the files. Sorry for the problems.
But am I right, that you now have the file correct?

[kar_bon]

Originally Posted by gd_barnes

I just now got the file from Karsten. Karsten, what compression software are you using? I'm still baffled.

i'm using not a really unzipper, i'm using TotalCommander V7.x. (great Filemanager -> http://www.ghisler.com/)
it has some packer/unpacker included so this tool can handle a zip/rar/arj/7z like a folder: insert/delete/copy from a compressed file as easy as can be!
unpacked Svens file (no problem) and packed again with this tool.

BTW: i'm using this for FTP-upload for my database, too!

PS: factors for 6500G-7000G sent to Gary!

[gd_barnes]

Originally Posted by Svenie25

I used WinZip 12.1 for this file, I think. Very strange that you become an error. For the next files I will use again the normal method and email you the files. Sorry for the problems.
But am I right, that you now have the file correct?

That is very strange. What would you consider the "normal" method? The latest version of WinRAR should be able to handle anything that WinZIP puts out.

Yes, Karsten sent me the file. Thanks for checking.

Compression software has become way overly complicated these days. I miss the days of only WinZIP.