Bases 251-500 reservations/statuses/primes

[rogue]

On the Sierpinski side I found:

3*322^1+1
4*322^1+1
6*322^1+1
7*322^2+1
9*322^2+1
10*322^1+1
12*322^4+1
13*322^2+1
15*322^1+1
16*322^1+1

k=1 is a GFN. I didn't do any testing. Are these to be tested separately? The other k have trivial factors.

On the Riesel side I found:

2*322^1-1
3*322^3-1
5*322^1-1
6*322^1-1
8*322^10-1
9*322^1-1
11*322^1-1
12*322^1-1
14*322^1-1
15*322^2-1
17*322^2-1

The other k have trivial factors. This conjecture is proven.

[gd_barnes]

Originally Posted by rogue

On the Sierpinski side I found:

k=1 is a GFN. I didn't do any testing. Are these to be tested separately? The other k have trivial factors.

You can choose to test them or not. I leave it up to the people that start the bases if they want to mess with testing GFNs since it's not necessary to prove or eventually prove the conjectures. If they don't, then I do.

I like to have these mentioned for historical reference only in case there is ever any change to the math that would allow much faster testing for GFNs at high n-limits.

After testing n<=2^15:
322^n+1 is a GFN with no known prime.


Gary

[appeldorff]

And I'm done xD
Sierp 255 is now at n=25000 with 267 k's remaining. Below is a file with all the primes (i sent you the primes up to n=15000 in case you forgot).

I have a zip file with the residues but it's far too big to attach to a forum post. How do you want it?

EDIT: Sent!

[gd_barnes]

Originally Posted by appeldorff

And I'm done xD
Sierp 255 is now at n=25000 with 267 k's remaining. Below is a file with all the primes (i sent you the primes up to n=15000 in case you forgot).

I have a zip file with the residues but it's far too big to attach to a forum post. How do you want it?

You can Email it to me at:

gbarnes017 at gmail dot com


Thanks!

[Xentar]

Originally Posted by gd_barnes

You can Email it to me at:

Do you really collect all these files? Why?

I've got a 30 MB file (5 MB compressed) für sierp b19...

[gd_barnes]

Originally Posted by Xentar

Do you really collect all these files? Why?

I've got a 30 MB file (5 MB compressed) für sierp b19...

Several reasons:

1. They've helped me detect missed ranges and k's. One that I can specifically remember: One person reported that xx base was complete to n=25K and posted the results. Upon looking in the file, I found that they missed searching one of the k-values.

2. When someone has a typo in a posted prime (I primality test ~half of all primes posted here as a double check), checking the results file for the correct prime saves a lot of time if the person can't mentally remember what the correct one is and it's not an obvious transposition of digits. If the person can't remember the correct one and doesn't have the results, then a lot of CPU time can be wasted double checking the range.

3. Matching up residuals vs. residuals on future double check efforts. Sometimes people might have bad memory (the computer kind, lol) and don't know it. I did one time on a work laptop. This causes bad residuals and hence will miss primes. If we find bad residuals in a particular area of a base, we can know to concentrate more on that area for missed primes and alert the person who searched it.

4. My own comfort level. Some people are extremely detailed and others are not. Some hang around for years and some come-and-go. For the less detailed types and people that come-and-go, it's not uncommon for them to simply miss posting a prime. A review of the results file sets my mind at ease in that regard.

The bottom line is that I try hard to avoid double-work on these bases. I'd hate to search a k up to n=1M and not find a prime and upon a double check, found that it had a prime at n=2000 because someone missed posting it within a myriad of 10's of other small primes. That's why, no matter who it is that starts a new base, I almost always run it up to to n=2500 to check it myself (unless it's a small proven conjecture). That's my small amount of double check to avoid a possible huge future search effort on various k's. On 2-3 occassions that I can recall, I've found primes completely missed by others for n<=2500. It's easy to miss 1-2 of them if you're manually typing them out of a file of 40-50 primes.

The above is why I much prefer that people post actual primes files instead of manually typing them into posts if they have more than ~5-6 primes to post for a base. I've found many many errors in primes manually typed into posts. If it's 1-2 larger primes, then it's no big deal but if it's quite a few smaller primes, posting a file is better. Computers don't transpose digits or mistype a k. :-)

I've had bigger than 5 MB compressed files sent to me, although it's best if they be < 10 MB compressed. (not sure how big google will allow but it's at least that big) If you have time, I'd really appreciate getting your file. What helps me the most is to get them in nice orderly ranges. If you can send me n=25K-30K for Riesel base 19, that'd be great.

That said, the most important results that I like to get are at the very high n-ranges, because there is much greater chance of a bad residual on a long test. (One bad internal calculation for any reason will make it an invalid test.) If you think it's too cumbersome to send me base 19, I'd much prefer to have results on your final k for base 17 (and for some other k's if you can). Those are the ones that we might want to cross-check the residuals on in a future double check effort.

My Email is gbarnes017 at gmail dot com.


Thanks,
Gary

[mdettweiler]

Originally Posted by gd_barnes

I've had bigger than 5 MB compressed files sent to me, although it's best if they be < 10 MB compressed. (not sure how big google will allow but it's at least that big) If you have time, I'd really appreciate getting your file. What helps me the most is to get them in nice orderly ranges. If you can send me n=25K-30K for Riesel base 19, that'd be great.

Yeah, I think Gmail's attachment limit is at least 10MB (coming or going), though if the sending account has a smaller limit on outgoing attachments the effective limit would be less. BTW, if anyone has a results file they need to get to Gary that's too big for email, drop me a line (max@noprimeleftbehind.net)--usually I should be able to come up with some way or other to get it where it needs to go.

[KEP]

Well according to Google Help, the maximum allowed size of an e-mail is: 20MB, so in the case you send an empty e-mail with just 1 attached file, then that file can be 20MB in size, before you send it. So to sum up, the maximum allowed attachement size is:

20MB-the actual size of your written e-mail

KEP

[mdettweiler]

Originally Posted by KEP

Well according to Google Help, the maximum allowed size of an e-mail is: 20MB, so in the case you send an empty e-mail with just 1 attached file, then that file can be 20MB in size, before you send it. So to sum up, the maximum allowed attachement size is:

20MB-the actual size of your written e-mail

KEP

Keep in mind that attachments must be encoded in base64 before transfer so that they can be packed into an all-ASCII email message; this has the effect of balooning the size a bit. For a huge file, that may add a MB or two.

[MyDogBuster]

Riesel Base 474
Conjectured k = 39
Covering Set = 5, 19
Trivial Factors k == 1 mod 11(11) and k == 1 mod 43(43)

Found Primes: 32k's File attached

Remaining k's:
4*474^n-1 <------ Proven composite by partial algebraic factors
9*474^n-1 <------ Proven composite by partial algebraic factors

Trivial Factor Eliminations:
12
23
34

Conjecture Proven

[gd_barnes]

Originally Posted by appeldorff

And I'm done xD
Sierp 255 is now at n=25000 with 267 k's remaining. Below is a file with all the primes (i sent you the primes up to n=15000 in case you forgot).

I have a zip file with the residues but it's far too big to attach to a forum post. How do you want it?

EDIT: Sent!

Appeldorf,

Per this post after your n=15K status, k=87036 with a prime at n=4784 was already eliminated. So there are now 266 k's remaining.

Thanks for the large effort!


Gary