[QUOTE=Chuck;526938]Looking at the available TF work, it looks like this person is taking all available candidates to 73 bits which have already been double-checked (clearing out the 38M to 43M ranges).[/QUOTE]

Yup. I can't go into too many details, but they specifically asked to have access to work which would help Wayne's "To below 2,000" sub-sub-sub-project. :tu:

Hey Chris, I know you are now with your nose up to the ears in the notebookz, but just to let you know that your computer doesn't know how to add. Proof is the fact that at least in the "Cost per Factor found in DC range" table, at least in the "73 bits" column, it says I only found 10 factors in total, where adding the cells above it by hand gets to at least 70 factors, so please send it to elementary school for few days so it can learn how to add.. :razz:

[QUOTE=LaurV;528696]Hey Chris, I know you are now with your nose up to the ears in the notebookz, but just to let you know that your computer doesn't know how to add.[/QUOTE]

Yeah, I know... Bad computer! Bad!

The sad thing is, of course, that it only does what I tell it to do... :wink:

A quick general update... A bunch of things are broken at the moment. This is because of a bit of data corruption in a single table.

The good news is I have backups, and this isn't affecting the assignments and "observations" of completeness.

The bad news is this table is a "living document" -- being updated several times a minute (sometimes several times a second).

I can't just import the backup. I have to very carefully have code identify the bad records, and then replace it with the (backed up) good data. Not code you run without testing and observation...

Plus, of course, I have "real work" to do as well. Gotta pay for these servers (and my kittens' feedings) somehow...

Would the above difficulties include the View Assignments page showing '2012-12-12' for the expected completion date for all TF work? This started a couple of days ago (not sure exactly when.) I'm currently 2507.00 days in arrears. :max:

[QUOTE=kladner;528768]Would the above difficulties include the View Assignments page showing '2012-12-12' for the expected completion date for all TF work? This started a couple of days ago (not sure exactly when.) I'm currently 2507.00 days in arrears. :max:[/QUOTE]

Yup... Again, sorry about this.

That date is something I set manually when I realized the table was (partially) corrupted. Note the pattern...

I've got a meeting with one of my "handlers" for another project later today, but I should be able to begin cleaning up this mess tonight.

It wouldn't be the end of the world, but please don't "unreserve" those assignments. They will magically disappear shortly...

[QUOTE=chalsall;528769]Yup... Again, sorry about this.

That date is something I set manually when I realized the table was (partially) corrupted. Note the pattern...

I've got a meeting with one of my "handlers" for another project later today, but I should be able to begin cleaning up this mess tonight.

It wouldn't be the end of the world, but please don't "unreserve" those assignments. They will magically disappear shortly...[/QUOTE]
No worries. All the other data displays correctly, including number of assignments. This just gave me a puzzle to pay attention to.

TF / P-1 status?
 

Chris, should the appearance of Ben Delo change our TF or P-1 strategy?

According to the Top Primality Producers page it appears Ben Delo is contributing 800000 GHz-days/month. About double curtisc's output. I assume this has greatly eaten into the reserve of available LL/PRP exponents that have been TF'ed and P-1'ed.

It's all about balancing TF resources with the LL/PRP resources. I'm wondering if it is time to reserve further ahead of the wavefront and do one bit less of TF. You are the one with the best handle on the capabilities of the TF'ers and prime testers.

If P-1 resources cannot keep up, then I should change PrimeNet to save a pool of P-1'ed exponents for manual reservations. Non P-1'ed exponents given to custisc, Ben Delo, or other prime95 users will get some P-1 done by the assignee.

[QUOTE=Prime95;529072]Chris, should the appearance of Ben Delo change our TF or P-1 strategy?[/QUOTE]

Not just Mr. Delo... There's also Ryan Propper to consider (while I appreciate he's doing a great deal of work, I do wish he'd complete some of the (tens) of thounds of P-1 assignments he currently holds in 93M and immediately above).

[QUOTE=Prime95;529072]According to the Top Primality Producers page it appears Ben Delo is contributing 800000 GHz-days/month. About double curtisc's output. I assume this has greatly eaten into the reserve of available LL/PRP exponents that have been TF'ed and P-1'ed.[/QUOTE]

I have been watching this myself, modeling how best to handle the situation.

Right now, the TF'ers have everything in [URL="https://www.mersenne.ca/status/tf/0/0/3/9000"]90M to 93M (inclusive) up to 77 bits[/URL] (I'm clearing the last 29 below that now). We're [URL="https://www.mersenne.ca/status/tf/0/0/4/9400"]about 65% of the way there for 94M[/URL].

~94.4M is, of course, where Primenet is currently handing out LL Cat 4.

[QUOTE=Prime95;529072]It's all about balancing TF resources with the LL/PRP resources.[/QUOTE]

Not just that -- we also have to include in the model the P-1'ers.

It isn't the end of the world when the P-1'ers overtake the TF'ers. The P-1'ers just spend a bit more time working each candidate. And, actually, it helps the TF'ers because it (slightly) lessens the work needed to be done.

But I would consider it "unfortunate" if Primenet ever handed out a candidate for LL work which had not had the TF'ing done "optionally", and the P-1 done "well".

This is particularily an issue for the P-1'ing -- Prime95/mprime only nominally use a lot of RAM when doing P-1. Might be a "suprise" to someone relatively new to the game.

[QUOTE=Prime95;529072]I'm wondering if it is time to reserve further ahead of the wavefront and do one bit less of TF. You are the one with the best handle on the capabilities of the TF'ers and prime testers.[/QUOTE]

It was getting to the time to pull in another 1M block of work, for the TF'ers. What I'll do instead is take two blocks -- 98 and 99M.

We do have some resources currently doing DCTF work who might be willing to move back to LLTF. I have myself moved all my GPUs (and all my "Instances") back to LLTF to 77, trying to keep our own P-1'ers appropriately fed.

I believe the TF'ers can adapt to the situation, and keep offering candidates TF'ed to at least 76 bits, and likely all the way to 77.

[QUOTE=Prime95;529072]If P-1 resources cannot keep up, then I should change PrimeNet to save a pool of P-1'ed exponents for manual reservations. Non P-1'ed exponents given to custisc, Ben Delo, or other prime95 users will get some P-1 done by the assignee.[/QUOTE]

A good plan. But...

One thing you might also want to look into are the current Manual Reservations for P-1 assignments in the lower 9xM ranges. There are a good 15,000 or so candidates already optimally TF'ed, but are "lagging" a bit (no joke intended) with the P-1 assignments...

Perhaps consider lower the allowed age for manually reserved P-1 assignments from six (6#) months to something more reasonable like two (2#) months.

Feedback, of course, welcome.

[QUOTE=chalsall;529076]Not just Mr. Delo... There's also Ryan Propper to consider (while I appreciate he's doing a great deal of work, I do wish he'd complete some of the (tens) of thounds of P-1 assignments he currently holds in 93M and immediately above).[/QUOTE]

I recall that when he started doing P-1 factoring a few months ago, he seemed to be only submitting known factors at first. So I think he completed the assignments, but maybe he had a broken script that wasn't collecting "no factor" results. Someone might want to contact him.

Email sent.

[QUOTE=ixfd64;529085]I recall that when he started doing P-1 factoring a few months ago, he seemed to be only submitting known factors at first. So I think he completed the assignments, but maybe he had a broken script that wasn't collecting "no factor" results. Someone might want to contact him.[/QUOTE]

Ryan's doing a TON of DC.; very little LL; As for P1 he's still doing them both with and without factors. Just not as many lately.

And even as fast as Ben is doing LL ... unless I'm reading it wrong aren't there are about 80,000 available in the 80-99M ranges?