HOWTO: Delete a k from dat file when a prime is found

[Xentar]

Reserving 2650 - 2700

ETA: 17 Nov.

Btw, when do we get a new *.dat file, with 273 k? :)

[michaf]

Originally Posted by Xentar

Reserving 2650 - 2700

ETA: 17 Nov.

Btw, when do we get a new *.dat file, with 273 k? :)

I can make a temporary file available, but we will miss out all new
factors in that one then :>.
Axn1 will get to it, someday...

For now, just use srfile -a -d on the datfile

[axn]

Originally Posted by michaf

I can make a temporary file available, but we will miss out all new
factors in that one then :>.
Axn1 will get to it, someday...

Hopefully, I will get to it by tomorrow this time. Since ltd is away, I have scheduled all my factor updates towards the weekend. Next week onwards, it will be back on schedule. Promise.

Originally Posted by michaf

For now, just use srfile -a -d on the datfile

That'd be sr5sieve -d <k>. Much easier!

[michaf]

Originally Posted by axn1

That'd be sr5sieve -d <k>. Much easier!

Hmm... I never knew that :> WAAAY easier...

[Xentar]

Originally Posted by axn1

That'd be sr5sieve -d <k>. Much easier!

This command deletes the k in both, Sierpinski and Riesel.
Is there any k, which is open in both projects?

(yeah - I know, it is Off-Topic )

Originally Posted by Xentar

Is there any k, which is open in both projects?

Obviously no.

[mdettweiler]

Originally Posted by Xentar

This command deletes the k in both, Sierpinski and Riesel.
Is there any k, which is open in both projects?

(yeah - I know, it is Off-Topic )

How do I delete just one Riesel k from my copy of the dat file? According to this thread, a new prime has been found on the Riesel side of the project, and since the new sieve file is not out yet, I'd like to remove the k from my dat file in the interim. However, if the command mentioned in this thread and others removes the k for both Sierpinski and Riesel, then what command do I need to enter to remove the Riesel k only?

[masser]

None of the Riesel k match Sierpinski k, so it is safe to use the command described in this thread. It probably will not affect the speed of srsieve.

[mdettweiler]

Originally Posted by masser

None of the Riesel k match Sierpinski k, so it is safe to use the command described in this thread. It probably will not affect the speed of srsieve.

Okay, thanks. But if it probably won't affect the speed, then why does anyone bother releasing new dat files?

Or, did you mean that it wouldn't affect srsieve, but it would affect sr5sieve? I use sr5sieve, FYI.

[masser]

That particular sequence will not remove a lot of candidates. Also, right now we have 228 sequences in the dat file, so in general each removed sequence does not affect the speed of the sieve that much.

However, after a couple of primes are found, their combined contribution to the dat file can be significant, so we try to update the dat file occasionally. Also, updating the dat file prevents submission of factors that are no longer needed. We also like to monitor our progress by tracking the size of the dat file - so that is another reason for providing updated dat files.

Hope this helps!

[mdettweiler]

Originally Posted by masser

That particular sequence will not remove a lot of candidates. Also, right now we have 228 sequences in the dat file, so in general each removed sequence does not affect the speed of the sieve that much.

However, after a couple of primes are found, their combined contribution to the dat file can be significant, so we try to update the dat file occasionally. Also, updating the dat file prevents submission of factors that are no longer needed. We also like to monitor our progress by tracking the size of the dat file - so that is another reason for providing updated dat files.

Hope this helps!

Ah, I see now. I removed the k anyway, since it's being run on an old computer for which every tiny speed increase counts. And, hey, I guess there's no reason not to remove the k even if it was on a faster computer--it's not like it's that hard to remove it, and it's never bad to have a speed increse.

When the official updated sieve file comes out, I'll use that instead, so that I can take advantage of all the individual factors removed as well.