Bases 6-32 reservations/statuses/primes

[gd_barnes]

k=2297 @ n=80K
k=2993 @ n=75K
k=13854 @ n=85K

[gd_barnes]

Status on my reservation for new Sierp base 19:
That is currently on hold. I think there were something like 2000+ k's remaining at n=3.5K and it was just crawling along! I'll pick that one up again in late Jan./early Feb.


Gary

[mdettweiler]

Originally Posted by gd_barnes

Status on my reservation for new Sierp base 19:
That is currently on hold. I think there were something like 2000+ k's remaining at n=3.5K and it was just crawling along! I'll pick that one up again in late Jan./early Feb.

You know, you could aways simply crunch it just enough to get it up to a more "round-numbered" n-value, then release it, and list all 2000 k's on their own page--then later, when you're ready to do more, you could reserve some of them if you wanted to. I guess it would take quite a while to make the web page for it, though.

BTW: I'm almost done sieving the 5 Sierp. Base 6 k's that I'd reserved. They should be done sometime around mid-day tomorrow. I've decided I'll keep two of them for PRP testing; I haven't decided which ones, yet.

[tnerual]

after some PM with Gary,

i will start sierpinski base 31 from scratch ... conjoncture is at 6360528 so i will have a lot of test to do

[Siemelink]

From the instructions:

(2) When starting a new base, you'll want to use PFGW using a script
Here is a cut-and-paste of one I used for base 30 Sierp:

ABC2 $a*30^$b+1 // {number_primes,$a,1}
a: from 1 to 866
b: from 1 to 5000

If I read this well, the script will still try the trivial k's. For b = 30 that's not too bad: k = = 28 mod 29. But for odd bases the trivial k == 1 mod 2 covers 50%. How do I get that in the script?

ABC2 $a*3^$b+1 // {number_primes,$a,1}
a: from 2 to 1000 step 2
b: from 1 to 5000

does not parse. I've poked around a bit but have no clue at the moment.
Anyone?

Willem.

[tnerual]

Originally Posted by Siemelink

From the instructions:

(2) When starting a new base, you'll want to use PFGW using a script
Here is a cut-and-paste of one I used for base 30 Sierp:

ABC2 $a*30^$b+1 // {number_primes,$a,1}
a: from 1 to 866
b: from 1 to 5000

If I read this well, the script will still try the trivial k's. For b = 30 that's not too bad: k = = 28 mod 29. But for odd bases the trivial k == 1 mod 2 covers 50%. How do I get that in the script?

ABC2 $a*3^$b+1 // {number_primes,$a,1}
a: from 2 to 1000 step 2
b: from 1 to 5000

does not parse. I've poked around a bit but have no clue at the moment.
Anyone?

Willem.

did you delete the pfgw.ini file ?

in fact, to start a new base, you need only 3 files in your pfgw directory (move all others to a "doc" directory)
the 3 files are

1. winpfgw.exe
2. install (not sure it's needed)
3. script.txt (your own script)

hope this help you

[michaf]

I just realised I only PM'd Gary about me starting the riesel side of base 24...

With a quite low conjecture it'll turn out to be a behemoth :>
432 candidates left at n=3300 :>

One comfort: it'll give quite a lot of primes above n=5k :)

[mdettweiler]

Sieving has been completed on my five reserved Sierp. Base 6 k's:
154797
157473
166753
168610
172257
I sieved up to p=600G for the range 30K<n<100K. I'm keeping 168610 and 172257 reserved for PRP testing; I'm releasing the rest. The sieve files for 154797, 157473, and 166753 are attached.

[gd_barnes]

Originally Posted by Siemelink

From the instructions:

(2) When starting a new base, you'll want to use PFGW using a script
Here is a cut-and-paste of one I used for base 30 Sierp:

ABC2 $a*30^$b+1 // {number_primes,$a,1}
a: from 1 to 866
b: from 1 to 5000

If I read this well, the script will still try the trivial k's. For b = 30 that's not too bad: k = = 28 mod 29. But for odd bases the trivial k == 1 mod 2 covers 50%. How do I get that in the script?

ABC2 $a*3^$b+1 // {number_primes,$a,1}
a: from 2 to 1000 step 2
b: from 1 to 5000

does not parse. I've poked around a bit but have no clue at the moment.
Anyone?

Willem.

You have to run multiple modulos for many bases. For instance, for Sierp base 31, you have to eliminate 1 mod 2, 2 mod 3 and 4 mod 5. Then it's just a matter of getting the LCM of 2, 3, and 5, i.e. 30, and eliminating the values that have at least one of the conditions. I had Tnueral run:

k == 0, 6, 10, 12, 16, 18, 22, and 28 mod 30

So k == 6 mod 30 for Sierp base 31 would be coded as:

ABC2 $a*31^$b+1 // {number_primes,$a,1}
a: from 6 to xxxxx step 30
b: from 1 to 5000

So, you would have 8 scripts to run for Sierp base 31.

I'm not sure what you meant by it 'does not parse'. The script that you show for base 3 should run just fine; only testing even candidates.

It's tricky when starting a new base. Between trivial and algebraic factors as well as multiples of the base, it's easy to end up testing far more than you need to. But not all MOB without a prime can be eliminated. Only the ones where k / base is still remaining.


Gary

[tnerual]

sierpinski base 31

all k tested up to n=580 ...
14326 k left

i continue on this case !

[gd_barnes]

Riesel base 16, k=2297, 2993, and 13854 complete to n=100K. No primes.

I am unreserving them.