Sierp base 16 - team drive #1

[Mini-Geek]

200K-210K is done. No primes. Results attached. I checked them myself, but a processing with the original sieve file would be appreciated.

[MyDogBuster]

Reserving 210K-220K

[MyDogBuster]

210-220 complete - nothing found

[gd_barnes]

A file has been posted for n=220K-250K. All k's with primes have been removed. In a manner similar to base 6, please cut your reservations out of the file in multiples of n=2K.

The file has been sieved for n=200K-500K and the sieve depth is P=12T. More sieving will be needed for n>250K.

[Mini-Geek]

Reserving 220K-230K.

[rogue]

Originally Posted by Mini-Geek

Reserving 220K-230K.

Wouldn't it be best to get a PRPNet server up that combines both Riesel and Sierpinski sides?

[Mini-Geek]

Originally Posted by rogue

Wouldn't it be best to get a PRPNet server up that combines both Riesel and Sierpinski sides?

Do you mean me, specifically, or in general? I've got a personal PRPnet server for my own work, (so I do have a PRPnet server up that combines Riesel and Sierp) but I could definitely see the use of a public PRPnet server searching R16 and S16 combined. Perhaps it'd be too much administrative overhead to try to match ranges from both sides and then separate results.

[rogue]

Originally Posted by Mini-Geek

Do you mean me, specifically, or in general? I've got a personal PRPnet server for my own work, (so I do have a PRPnet server up that combines Riesel and Sierp) but I could definitely see the use of a public PRPnet server searching R16 and S16 combined. Perhaps it'd be too much administrative overhead to try to match ranges from both sides and then separate results.

A public one of course. They have one for NPLB, why not put a CRUS one that the same server.

Separating results would be easy, presuming one knows a little SQL.

[mdettweiler]

Originally Posted by Mini-Geek

Do you mean me, specifically, or in general? I've got a personal PRPnet server for my own work, (so I do have a PRPnet server up that combines Riesel and Sierp) but I could definitely see the use of a public PRPnet server searching R16 and S16 combined. Perhaps it'd be too much administrative overhead to try to match ranges from both sides and then separate results.

Originally Posted by rogue

A public one of course. They have one for NPLB, why not put a CRUS one that the same server.

Separating results would be easy, presuming one knows a little SQL.

Indeed, I believe this has in fact been suggested before (over in the R31 PRPnet mini-drive thread, IIRC). If the interest is great enough, then I can quite easily set up a PRPnet server for base 16.

Regarding separating results, that's no problem. I don't actually use SQL to do it (though that would be another good way to do it); rather what I do is run the completed_tests.log files through my suite of results processing scripts, which includes a script for pulling out subsets of results files. In fact, I did exactly this not long ago for Gary's personal PRPnet server when he ran R28 and S28 simultaneously.

[gd_barnes]

I'd like to leave this open to the less-online folks also. Let's get both drives up to n=250K, sieve to an optimum depth for n=250K-500K, and then load a moderate-sized range into a PRPnet server to see what the response is. The response has been quite tepid for the PRPnet mini-drive that has included various bases in the 20s and now base 31 at sizes mostly comparable to this one.

[rogue]

Originally Posted by gd_barnes

I'd like to leave this open to the less-online folks also. Let's get both drives up to n=250K, sieve to an optimum depth for n=250K-500K, and then load a moderate-sized range into a PRPnet server to see what the response is. The response has been quite tepid for the PRPnet mini-drive that has included various bases in the 20s and now base 31 at sizes mostly comparable to this one.

I can understand that response would be tepid. There are not a lot of people who are interested in the CRUS project. I suspect that PrimeGrid would have no issue if someone from mersenneforum would promote a PRPNet server for CRUS over on their forum. It has already been done for NPLB and I think that the possibility of either proving a conjecture or finding a Top 5000 prime would bring a few other searchers over to CRUS.