[QUOTE=chalsall;403763]As explained elsewhere, George asked me to bring in those assignments not yet appropriately TF'ed. Because of the way I do this, any candidates assigned by a non registered user via the Primenet Manual Assignment form is transferred to GPU Factoring.

I consider this a (Primenet) bug; why on god's green earth should an unregistered user be allowed to reserve work? (Hint: someone malicious could easily cause havoc.)[/QUOTE]

Ah, that explains it. I did think it was unusual. That arrangement must be something specific to the GPU72 reservation system, or something like that.

Still, that would mean that if we wanted to unreserve those from LL work, since they belong to that account you should be able to unreserve and then check 'em out as TF assignments without too much trouble, I think? I mean, if that was the goal... to free up those LL assignments. 802 could be wiped off the board in one swell foop.

As for the rest, it's really George's call as far as the expirations go. The grandfather rules do stipulate that many of them *should* be expired already, they've just been granted a reprieve since it'll be years (decades?) before those exponents are being passed out on a regular basis. If it made sense, George could adjust that expiration task.

For grandfathered assignments, even if it meets all the other criteria (old assignment, outside the "grace period" for work in progress), it will ignore it if it's above the daily "critical threshold".

Today's current critical threshold for first time LL checks is: 59288870 Anything above that is ignored for the purposes of expiring grandfathered work. If it weren't for that exception, most of those would have expired already since they only get a 1-year time period for work that hasn't even started yet.

There's also a general condition that applies to grandfathered and newer assignments alike: Don't expire them if they're more than 40M past a certain "base" exponent (currently 51M). So there again, anything over 91M would essentially be ignored.

On the other hand, expiring some of these that actually *are* progressing, albeit slowly, is a danger. These do take a very long time to complete, and if it started looking at expiring past the critical threshold, there's a high chance it would catch some of these and start expiring them even though they've been chugging along slowly but surely.

I just did a trial check and removing that "less than critical threshold" would result in the expiration of a number of assignments that are less than 1% away from finishing, like M332284963. It just happens to have been assigned long enough ago that it's past it's 665 day "hard expiration" period.

What George *could* do is modify it some so it's looking at "exponent < critical threshold *OR* % done = 0.0". That would ensnare the ones that haven't started yet, and save the ones that are actually doing *something* (without regard to if the % done is TF, P-1, LL, etc... at least it's doing something). Or maybe throw another OR in there that looks to see if the assignment has reported in during the past year. Is a year too much to ask for the client to say "howdy, I'm alive"? :smile:

[QUOTE=Madpoo;403779]Still, that would mean that if we wanted to unreserve those from LL work, since they belong to that account you should be able to unreserve and then check 'em out as TF assignments without too much trouble, I think? I mean, if that was the goal... to free up those LL assignments. 802 could be wiped off the board in one swell foop.[/QUOTE]

Underway.

Two instructions to my spiders.

[QUOTE=chalsall;403780]Underway.

Two instructions to my spiders.[/QUOTE]

Boom, just like that, now there are only 1348 of those grandfathered assignments > 332M with zero progress. :smile:
EDIT: make that 1169... I forgot to filter just on "progress=0"

Currently GPU Computing still owns 89 assignments in that category, for anyone doing the math to see how the #'s changed.

@Uncwilly: let me know if that means you need a new list of what's left?

[QUOTE=Madpoo;403781]@Uncwilly: let me know if that means you need a new list of what's left?[/QUOTE]
Yes, please could you? I don't want to cross paths with GPU72. :boxer:

I d/l'ed the old file.

[QUOTE=Uncwilly;403785]Yes, please could you? I don't want to cross paths with GPU72. :boxer:[/QUOTE]

There's an old saying: I fight with my friends because it's fun.

But, if you fight my friends, be prepared for a serious fight.

[QUOTE=chalsall;403791]There's an old saying: I fight with my friends because it's fun.

But, if you fight my friends, be prepared for a serious fight.[/QUOTE]

Is this a private fight, or can anyone join in? :smile:

For UncWilly: I've updated that file for you, in the same location as before.

Hopefully you can run some of them to higher bit levels or whatever and take care of these old timey assignments by finding factors for them.

[QUOTE=Madpoo;403796]For UncWilly: I've updated that file for you, in the same location as before.[/QUOTE]
[YOUTUBE]_crvGziPn3g[/YOUTUBE]

Here's #1. Three more coming within 2 weeks. (Everything else is using 2 core and shooting for long-term goals)

usslcburnin46 332406131 C - Unverified 2015-06-11 07:03 90.4 8203553C92E781__ 5240.5904

[QUOTE=Madpoo;403796]IHopefully you can run some of them to higher bit levels or whatever and take care of these old timey assignments by finding factors for them.[/QUOTE]
Here are some that I have done like that.

[url]http://www.mersenne.org/report_exponent/?exp_lo=332282569&full=1[/url]
[url]http://www.mersenne.org/report_exponent/?exp_lo=332267701&full=1[/url]
[url]http://www.mersenne.org/report_exponent/?exp_lo=332268329&full=1[/url]
[url]http://www.mersenne.org/report_exponent/?exp_lo=332281787&full=1[/url]
[url]http://www.mersenne.org/report_exponent/?exp_lo=332360239&full=1[/url]
[url]http://www.mersenne.org/report_exponent/?exp_lo=332364817&full=1[/url]
[url]http://www.mersenne.org/report_exponent/?exp_lo=332365871&full=1[/url]
[url]http://www.mersenne.org/report_exponent/?exp_lo=332367403&full=1[/url]
[url]http://www.mersenne.org/report_exponent/?exp_lo=332368403&full=1[/url]

[url]http://www.mersenne.org/report_exponent/?exp_lo=332249051&full=1[/url]
[url]http://www.mersenne.org/report_exponent/?exp_lo=332255177&full=1[/url]
[url]http://www.mersenne.org/report_exponent/?exp_lo=332284973&full=1[/url]
[url]http://www.mersenne.org/report_exponent/?exp_lo=332293303&full=1[/url]

[QUOTE=Uncwilly;402582]I noticed that you got a nice P-1 factor in the 91 or 92 bit range. :)[/QUOTE]
[QUOTE=aurashift;402588]That wasn't on a RAMBeast. That's why I was getting a little irked, that was 8,10, or 12 cores that had a bunch of wasted time.....22 days. Just looked it up.[/QUOTE]
I noticed several in the 80 bit range that you found. :tu:

[QUOTE=Uncwilly;404379]I noticed several in the 80 bit range that you found. :tu:[/QUOTE]

Isn't it awful? I hope they add 100M P-1 factoring to a future P95 version.

That's part of why I decided to switch everything that has AVX over to 2 core 100M workers.