Doublecheck drive #1: k=3-1001 n=100K-260K

[gd_barnes]

Originally Posted by Anonymous

Hmm...I was trying to make sense of Prof. Keller's k<300 page just now, but it's kinda hard to follow--I can't even locate the list of double-checked ranges at all, just the first-pass ones. I'm sure they're hiding in there somewhere, though; could you please point me to them?

The page is all first-pass ranges. There are no double-check ranges.

THAT is why I said RPS has dropped the ball on reporting their ranges. The ranges that we searched should have been reported long ago. It's part of the reason why I wanted to do this drive. It'll be cool when all the ranges show as 'officially' searched up to n=260K on the page!

That reminds me, I need to report my k=37, 41, 43, 61, and 63 ranges as having been searched. I can't remember the exact ranges that I searched for them. It seems like it was n=600K-1M for all of them and perhaps n=500K-1M for one of them. I need to cull my results off of multiple machines to remember for sure.


Gary

[gd_barnes]

Max,

Please resort the files for this drive by k-value instead of n-value. Two reasons:

1) It will be FAR easier to visually double-check all primes by k-value against Karsten's site and Primesearch. Nowhere that I know of is a list of primes in this range by n-value.

2) It will allow people to reserve 1 k at a time. As it is, the 3k files are a bit big for some people's tastes and it would be real annoying to reserve a partial file with 3 k's for n=100K-150K or 100K-200K and leave the rest of it for someone else.

I'm asking now because right after the 1st drive is done, I'm going to take k=27-31 and will be taking quite a few files on-and-off in 2009 in order to promote the effort a little more.


Thanks,
Gary

[mdettweiler]

Originally Posted by gd_barnes

Max,

Please resort the files for this drive by k-value instead of n-value. Two reasons:

1) It will be FAR easier to visually double-check all primes by k-value against Karsten's site and Primesearch. Nowhere that I know of is a list of primes in this range by n-value.

2) It will allow people to reserve 1 k at a time. As it is, the 3k files are a bit big for some people's tastes and it would be real annoying to reserve a partial file with 3 k's for n=100K-150K or 100K-200K and leave the rest of it for someone else.

I'm asking now because right after the 1st drive is done, I'm going to take k=27-31 and will be taking quite a few files on-and-off in 2009 in order to promote the effort a little more.


Thanks,
Gary

Okay, I'll post the files for 1 k at a time later today. (Not sure exactly when later today, though, since I'll be out of the house for part of the day.) Feel free to bug me if it's not done by tomorrow.

[mdettweiler]

Originally Posted by gd_barnes

Max,

Please resort the files for this drive by k-value instead of n-value. Two reasons:

1) It will be FAR easier to visually double-check all primes by k-value against Karsten's site and Primesearch. Nowhere that I know of is a list of primes in this range by n-value.

2) It will allow people to reserve 1 k at a time. As it is, the 3k files are a bit big for some people's tastes and it would be real annoying to reserve a partial file with 3 k's for n=100K-150K or 100K-200K and leave the rest of it for someone else.

I'm asking now because right after the 1st drive is done, I'm going to take k=27-31 and will be taking quite a few files on-and-off in 2009 in order to promote the effort a little more.


Thanks,
Gary

Okay, should be all set now.

I've also changed the major sub-ranges around so that it's split up as k=3-300 and 300-1001, instead of 3-400 and 400-1001 as we did previously. Now it's a bit more consistent with the way we split up k-ranges for first-pass testing.

[gd_barnes]

Originally Posted by mdettweiler

Okay, should be all set now.

I've also changed the major sub-ranges around so that it's split up as k=3-300 and 300-1001, instead of 3-400 and 400-1001 as we did previously. Now it's a bit more consistent with the way we split up k-ranges for first-pass testing.

You read my mind! Thanks a bunch for taking care of that!

Actually, my idea was to leave them as files that contained 3 k's at a time and sorted by k-value but to allow 1k reservations. But I see that I wasn't specific in that regard. Alas, no biggie. The way it is works fine.

I'll reserve k=27 thru 31 now to start testing after the 1st drive is done.


Gary

[Flatlander]

Taking k=33.

(I've stress-tested with P95, cleared dust from the inside of the PC and lowered the overclocking slightly.)

[mdettweiler]

Originally Posted by Flatlander

Taking k=33.

(I've stress-tested with P95, cleared dust from the inside of the PC and lowered the overclocking slightly.)

Nice to see this effort picking up some speed again!

I guess I should probably post some more files sometime soon for the k<300 range...

[gd_barnes]

Originally Posted by mdettweiler

Nice to see this effort picking up some speed again!

I guess I should probably post some more files sometime soon for the k<300 range...

Or I could post some files.

My 3 k's are just now starting. I'm going to test k=31 first. It's the k that had the 2 missing top-5000 primes reported in the last couple of weeks. I'll be curious to see if everything is OK with this range.

After completing these 3 k's, I think I'll double-check k=31 up to n=500K where we know that Steve at RPS started testing it. I see that Karsten shows a "?" by the 500K search range at rieselprime.org. Doing the double check that high should eliminate it.

I don't like those question marks. They're kind of like gaps! There's another one of my sayings for your book, Chris. "I hate question marks." lmao


Gary

[mdettweiler]

Originally Posted by gd_barnes

Or I could post some files.

My 3 k's are just now starting. I'm going to test k=31 first. It's the k that had the 2 missing top-5000 primes reported in the last couple of weeks. I'll be curious to see if everything is OK with this range.

After completing these 3 k's, I think I'll double-check k=31 up to n=500K where we know that Steve at RPS started testing it. I see that Karsten shows a "?" by the 500K search range at rieselprime.org. Doing the double check that high should eliminate it.

I don't like those question marks. They're kind of like gaps! There's another one of my sayings for your book, Chris. "I hate question marks." lmao


Gary

Okay, cool--I forgot that you have access to the doublecheck drive sieve file too.

As for doublechecking k=31: okay, that sounds good. When we do our next doublecheck range after this drive is complete, make sure you remind me to skip re-doublechecking that range if I don't remember already.

[Flatlander]

k=33 complete.
No surprises:

33 109848
33 128635
33 136971
33 173110
33 174182

[gd_barnes]

Originally Posted by Flatlander

k=33 complete.
No surprises:

33 109848
33 128635
33 136971
33 173110
33 174182

Chris,

I don't know if you are aware of this but you'll need to report your range completed to Wilfred Keller's page per post 48 in this thread.

It just so happens that the page only has k=33 as reported to n=250K. Since you searched it to n=260K, then it needs to be reported there.

There should be enough detailed instructions in post 48 for you to take care of it and it shouldn't take more than 5-10 mins. Let us know if there are any questions.

For my 3 k's, they are all already reported there past n=260K so I won't have to report any of them when I'm done.

Status: k=31 is nearing completion now and k=29 will start shortly. No problems found yet. For the next several months except during rallies, I'm going to keep one core running constantly on this effort.


Gary