Does anybody want to take a shot at [url=http://www.mersenne.org/report_exponent/?exp_lo=79299719&full=1]79299719[/url]?

[QUOTE=TObject;401144]Does anybody want to take a shot at [url=http://www.mersenne.org/report_exponent/?exp_lo=79299719&full=1]79299719[/url]?[/QUOTE]

Sure, why not. I might as well actually do some LL on occasion lol

[QUOTE=Mark Rose;401147]Sure, why not. I might as well actually do some LL on occasion lol[/QUOTE]

I hope the 4th time is the charm. Holler if it needs a 5th check. :smile:

[QUOTE=Madpoo;401166]I hope the 4th time is the charm. Holler if it needs a 5th check. :smile:[/QUOTE]

I'm running it on an EC2 instance with a few extra cores with absolutely nothing else to do. It says it'll take about three weeks.

[QUOTE=Mark Rose;401176]I'm running it on an EC2 instance with a few extra cores with absolutely nothing else to do. It says it'll take about three weeks.[/QUOTE]

Please let us know who's residue you match with (presuming you do). I don't know why I'm listed as one of the attempts -- it didn't pass through the GPU72 proxy....

[QUOTE=chalsall;401245]Please let us know who's residue you match with (presuming you do). I don't know why I'm listed as one of the attempts -- it didn't pass through the GPU72 proxy....[/QUOTE]

Will do!

Last 3 of the 54M first-time checks
 

I gotta be honest folks, the original assignees of these 3 exponents will not finish them in time:
54357769
54674791
54759797

The first one, despite being 97.9% done, is moving along at 0.0053% per day, with a real ETA of 2016-06-02.

The other 2 haven't checked in at all since April 15th. They might surprise me and check in again sometime soon, but from their previous rate of progress they're still only estimated to finish in March 2016.

All 3 would have expired before then anyway.

The remaining 55M exponents look better, with 55771997 as the exception (projected ETA in Feb 2016). The rest of those might squeak in around May/June of this year.

As near as I can tell, there's no provision for grandfathered assignments to expire after XX days of no contact. They'll only expire once they reach that max of 665 days (assuming they were at 99.9% complete by then).

With the last two 54M numbers only being 67 / 66.2 % done, and taking into account their rate of progress of < 0.01% per day, they'll actually expire around the end of May for 54759797 and the first week of June for 54674791.

I'd estimate 54357769 as expiring closer to it's 665 day max (which is June 17, 2015). By just a couple days... maybe June 15.

It kind of boils down to when they check in, but that's it more or less.

On the other hand, if those last two continue to NOT check in, and their current % done remains unchanged, they'll expire May 25 and May 22 respectively.

In summary, here's my projected expirations (I assume the last two will NOT check in again with any progress beforehand):
54357769 - June 15
54674791 - May 25
54759797 - May 22

They should get reassigned pretty quick... I have my own residues for these ready to check in but I'll do those as double-checks.

[QUOTE=Madpoo;401488]
They should get reassigned pretty quick... I have my own residues for these ready to check in but I'll do those as double-checks.[/QUOTE]

No reason to do that IMO. Just submit your results the day before they would expire. (And perhaps manually verify with the actual expiration that they will indeed expire... or perhaps not.)

We should still continue to assume that these people will complete their test eventually some day, and they can be the double check, rather than wasting their work be re-assigning when we already have a second test complete.

[QUOTE=Dubslow;401500]No reason to do that IMO. Just submit your results the day before they would expire. (And perhaps manually verify with the actual expiration that they will indeed expire... or perhaps not.)

We should still continue to assume that these people will complete their test eventually some day, and they can be the double check, rather than wasting their work be re-assigning when we already have a second test complete.[/QUOTE]

I could do that I suppose.

To come up with the probably expiration dates, I used the same code that it look at to see if it's past the grace period or not (for % of work done), and then also checked to see if the % done would be less-than or greater-than that "grace %" on whatever date.

It was a little tedious and I could have actually done the math, but it was easier for me to just plug in different values in something like this:
(10 + (DATEDIFF (DAY, a.dt_when_assigned, getdate()) - 365) / 3.33)

As in, instead of "getdate()" I'd plug in something like '2015-05-31' and see what the grace % would be on that day, and then using my observed %/day increase see where it would be on that same day. If it's less than the grace % would be on that day, boom, expired.

It's a little fuzzy because these clients don't check in each day, and even if they did it might be unchanged from the last one since it goes in 0.1% increments. Which can take a couple weeks for some of the slowpokes.

That means they might expire sooner. Now, in the last 2 I assumed they wouldn't check in again. The first one has been steadily reporting, just not usually incrementing the % done. We'll see how that goes.

Well, Danny's been poached upon again. At least he still has those 2 poached assignments as DC's.

7 exponents now are left for the sub-56M milestone.

[QUOTE=cuBerBruce;401890]Well, Danny's been poached upon again. At least he still has those 2 poached assignments as DC's.

7 exponents now are left for the sub-56M milestone.[/QUOTE]

[URL="http://www.mersenne.org/M55059383"]M55059383[/URL]

Bummer... I was predicting that one to finish around June 15 (the other two assigned to him I'm predicting June 14 and June 19.

Yeah, when it checks in it'll be a double-check. Good enough I guess.

[QUOTE=cuBerBruce;401890]Well, Danny's been poached upon again. At least he still has those 2 poached assignments as DC's.

7 exponents now are left for the sub-56M milestone.[/QUOTE]

And only 15 left sub-57M. Sub-66M might be the next logical countdown: 8,928 untested exponents remain.

[QUOTE=NBtarheel_33;401927]And only 15 left sub-57M. Sub-66M might be the next logical countdown: 8,928 untested exponents remain.[/QUOTE]

Also, the assignments for 54674791 and 54759797 checked in today... not much different than when they last checked in and the prognosis is even worse, with my own projections of finishing by early in the year 2020. At the current rate they'll expire anyway in just a couple weeks.

The owner of M55861261 has finally completed that assignment after 510.6 days. (Hurray!)

There are now 6 sub-56M assignments (officially) remaining for that milestone.

We'll see if any of the others will be finished before expiring (and hopefully, without poachers interfering). At least two of them appear to be on track to finish before being expired.

[QUOTE=cuBerBruce;402090]The owner of M55861261 has finally completed that assignment after 510.6 days. (Hurray!)

There are now 6 sub-56M assignments (officially) remaining for that milestone.

We'll see if any of the others will be finished before expiring (and hopefully, without poachers interfering). At least two of them appear to be on track to finish before being expired.[/QUOTE]

Here's my latest predictions:
[CODE]exponent RealEta DaysToHardExpire
54357769 2015-11-03 22:54 30.4
54674791 2020-01-05 05:06 14.5
54759797 2020-02-15 18:05 11.8
55027163 2015-06-17 18:18 65.5
55079077 2015-06-21 16:30 62.8
55771997 2016-03-04 02:38 111.5[/CODE]

The "DaysToHardExpire" is how long before it would expire if the client stopped reporting any progress at all from right now, according to the grandfather rules.

[QUOTE=Madpoo;402110]Here's my latest predictions:
[CODE]exponent RealEta DaysToHardExpire
54357769 2015-11-03 22:54 30.4
[/CODE]
[/QUOTE]

Wow, user "Ollum98" must be eating his Wheaties because the assignment for M54357769 is showing a big burst of speed. It had been poking along at something like 0.1% per week (per week!) and then just in the past 3 days it's shot up by 0.8%.

That obviously skews the 30-day trend of it's progress so I can't make a solid prediction. If it can keep up the latest rate of progress though, it *should* finish in just a couple more days since it's at 98.9% right now.

[QUOTE=Madpoo;402263]Wow, user "Ollum98" must be eating his Wheaties because the assignment for M54357769 is showing a big burst of speed. It had been poking along at something like 0.1% per week (per week!) and then just in the past 3 days it's shot up by 0.8%.

That obviously skews the 30-day trend of it's progress so I can't make a solid prediction. If it can keep up the latest rate of progress though, it *should* finish in just a couple more days since it's at 98.9% right now.[/QUOTE]

Now up to 99.0%. Perhaps this person has recently realized that it may expire soon, and has started giving it more runtime than usual.

[QUOTE=chalsall;401245]Please let us know who's residue you match with (presuming you do). I don't know why I'm listed as one of the attempts -- it didn't pass through the GPU72 proxy....[/QUOTE]

M79299719 is not prime.

[url=http://www.mersenne.org/report_exponent/?exp_lo=79299719&full=1]Bad news[/url] for For Research.

[QUOTE=Mark Rose;402303][url=http://www.mersenne.org/report_exponent/?exp_lo=79299719&full=1]Bad news[/url] for For Research.[/QUOTE]

Hmmmm... Thanks for the update.

Must investigate....

I just poached 2 of the last 3 54M exponents
 

I just checked in my own results for:
M54674791
M54759797

They were down to the point where the original assignment had less than a week left to finish it up before it expired at the rate it was going.

My latest projection on those was a completion date of Nov/Dec 2018.

So, now they're done with a first time check. Those assignments are converted to a double-check and maybe, just maybe, they'll finish up before they get reassigned to someone else as a double-check.

That leaves just M54357769 as the lone 54M first-time check. It's at 99%, and with it's recent "growth spurt" it may just finish in time. It is funny though... it went through a spurt there where it went up a whole 1% in 10 days, but then hasn't moved at all for the past 3 days... stuck at 99%. It might be back to it's slow progress.

Whatever the case, it'll expire irregardless in 29-30 days at most even if it managed to jump up to 99.9% by then. Or 26 days if it's stuck at 99%. We'll know a little better as we get closer to that day... if it doesn't seem like it'll make it I'll check in my result for it as well so at least the original assignee will be converted to a double-check. I already know it's not prime (oops... "spoiler alert").

[QUOTE=Madpoo;402518] irregardless[/QUOTE]

:bob:

[QUOTE]...[URL="http://en.wikipedia.org/wiki/Irregardless"]irregardless[/URL]...[/QUOTE]

Don't you judge me! :smile: I've been known to let an "ain't" slip into my vocabulary... some folks would pass out.

The mere fact that you still knew what I meant means that usage is fine. And that's how the English language has become such a mess, thanks to sloppy native speakers like myself. LOL

[QUOTE=Madpoo;402520]Don't you judge me! :smile: I've been known to let an "ain't" slip into my vocabulary... some folks would pass out.[/QUOTE]Shirley for all intensive porpoises without further or due your gonna stay cool, calm and collective under pressure, rite?

[QUOTE=retina;402521]Shirley for all intensive porpoises without further or due your gonna stay cool, calm and collective under pressure, rite?[/QUOTE]

A'ight!

Countdown to first time checking all exponents below 57M: 4
 

The text is lying. Or the count is lying. Both can't be right.

Either: Countdown to first time checking all exponents below 57M: 8
Or: Countdown to first time checking all exponents between 56M and 57M: 4

[QUOTE=retina;402570]The text is lying. Or the count is lying. Both can't be right.

Either: Countdown to first time checking all exponents below 57M: 8
Or: Countdown to first time checking all exponents between 56M and 57M: 4[/QUOTE]

The countdown lies.

I cheated a little bit... in the countdown for < 57M it's *really* the countdown for stuff between 56M and 57M. That's because when the <56M is eventually done, then that will be correct.

Anyway, it's wrong, but deliberately so, based on the SQL query used. I could make the wording more clear, or just go with a total <57M that includes the remaining 54M/55M work, but those won't be there too much longer.

Ditto on the double-check milestones. Call me lazy. :) If you click on the links to see the assignment report, it only includes the stuff for that specific range.

All exponents below [B]33,776,779[/B] have been tested and double-checked.
All exponents below [B][COLOR="Blue"]54,357,769[/COLOR][/B] have been tested at least once.

Countdown to testing all exponents below M([B][COLOR="Blue"]57885161[/COLOR][/B]) once: 916
Countdown to first time checking all exponents below 56M: [B][COLOR="Red"]4[/COLOR][/B] (Estimated completion : [COLOR="Green"]2015-07-05[/COLOR])
Countdown to first time checking all exponents below 57M: [B][COLOR="Red"]4[/COLOR][/B] (Estimated completion : [COLOR="Green"]2015-06-04[/COLOR])
Countdown to first time checking all exponents below 58M: [B][COLOR="Red"]908[/COLOR][/B] (140 still unassigned)

Countdown to double checking all exponents below 34M: [B][COLOR="Red"]5[/COLOR][/B] (Estimated completion : [COLOR="Green"]2015-05-28[/COLOR])
Countdown to double checking all exponents below 35M: [B][COLOR="Red"]9,380[/COLOR][/B] (8,630 still unassigned)

Countdown to proving M([COLOR="Green"]37156667[/COLOR]) is the [COLOR="green"]45[/COLOR]th Mersenne Prime: 40,175

[QUOTE=Madpoo;402580]Ditto on the double-check milestones. Call me lazy. :) If you click on the links to see the assignment report, it only includes the stuff for that specific range.[/QUOTE]

Specifically for the milestones with very few remaining exponents...

The lone 54M exponent should finish (I hope) by June 15 but as I mentioned earlier, it's progress has been pretty sporadic. It will definitely expire no matter what once it reaches it's 665 day old limit in 29 more days at the most. If it fails to progress at all past 99% it will expire in just 25 days. I'll check in my own result for that one once the assigned user checks theirs in, or just before it would expire anyway.

For the 55M exponents (3 of them), one of those probably won't finish until March 2016, but it'll expire before then (104 days if it didn't progress at all, but a bit longer if it continues on it's plodding rate of progress). The other two should finish around the end of June or early July.

On the 56M exponents (4 of them), all but one of those shows a pretty accurate estimate right there on the report page. One of them has an ETA of June 4th but I'm estimating June 17. Close enough. They'll all make it.

For the double-check milestone, there was one that checked in today (it had expired and the new assignee finished it in a day). Of the 5 left, none will finish before they're expired.

Those 5 are NOT grandfathered, so different rules apply. I checked out how the rules apply to those, and they're in a category that gives them 240 days to finish. One only has 11 days left before that, and at it's current rate it'll take another 104 days. The other 4 have between 50-60 days before they expire but they're going to take 230-240 more days to finish, so they'll expire for sure in a couple months.

I hope that helps explain why I didn't try that hard to clarify the "how many are left" #'s since they'll only be there a couple months at most. LOL

(I probably spent more time writing this than it would have taken to just clarify the language or correct the query). :smile:

[QUOTE=Madpoo;402518]
(I probably spent more time writing this than it would have taken to just clarify the language or correct the query). :smile:[/QUOTE]

You've explained only for this particular few months of time. But in a year the code will still be inaccurate, and you'll have to explain how we simply need to wait a few months again.

Best just fix the query now and be done with it.

Countdown to testing all exponents below M(57885161) once: 911
Countdown to first time checking all exponents below 58M: 903 (135 still unassigned)

Doesn't add up in my interpretation?

No point in having a countdown that is deliberately wrong., just remove it or fix it in my opinion. Only those of us following this thread have any idea about when exponents will expire/finish and it makes no sense people would have to know that in order to find the correct value of the countdown.

[QUOTE=ATH;402603]No point in having a countdown that is deliberately wrong., just remove it or fix it in my opinion. Only those of us following this thread have any idea about when exponents will expire/finish and it makes no sense people would have to know that in order to find the correct value of the countdown.[/QUOTE]

There, fixed. :smile:

EDIT: FYI, I really was just trying to get those new stats out there sooner. I think in the past, the new milestones only showed up once the current one had been checked off. I had those changes ready to go and I was double-checking the thing that shows how many are unassigned, and figured I'd have those show up on the actual page. Seems like it was just more confusing than helpful. Whoops.

< 34M double-checked
 

I just checked in results for the last 5 exponents in the 33M range.

See my analysis above, but in summary, one of them only had 10 more days to finish before it expired but it wouldn't have finished for another 104 days. The other 4 had 50-60 days to finish but were going to take another 225-240 days, so they would have expired before finishing as well.

Now that I have a setup to track the actual progress, I feel more confident about the decision to poach them.

I also have some insight on why this particular user's work might be going so darn slow. The user in question is running all of this work on a 6-core E5-1650 processor with one assignment on each of the 6 cores.

I know from my own experimentation that even 2 assignments in that range will seriously flood the memory channel and slow down both of them. I can only speculate what kind of thrashing goes on with *six* of those going on at the same time on one CPU. No wonder it was only moving along at 0.3% per day on average for each one.

If the user had merely setup all 6 cores in a single worker, it would probably take 36-40 hours per exponent in that 33M range and all 6 of those would have been done in a couple weeks. The way it is now, it'd be lucky to finish 6 assignments in under a year.

Anyway, milestone page is duly updated.

[QUOTE=Madpoo;402616]...Anyway, milestone page is duly updated.[/QUOTE]

Umm... before anyone points this out:

The milestone page shows 897 left for the < 58M, with 116 unassigned (as of right now). By my math that should mean there are 781 assignments for first-time checks in that range.

However, if you actually look at the # of assignments here:
[URL="http://www.mersenne.org/assignments/?exp_lo=54000000&exp_hi=58000000&execm=1&exp1=1&extf=1&exdchk=1"]http://www.mersenne.org/assignments/?exp_lo=54000000&exp_hi=58000000&execm=1&exp1=1&extf=1&exdchk=1[/URL]

You'll count just 778. Where are those other 3?

I wondered too so I dug into it. For some reason there are 3 exponents where the first check is marked as "suspect" and they currently aren't reassigned to anyone. For whatever reason it's messing up the query counts. I'd actually noticed the discrepancy before but haven't had time to dig into it. Now that I know whassup [sic] I might be able to fix that and get it accurate again.

[QUOTE=Madpoo;402637]...For some reason there are 3 exponents where the first check is marked as "suspect" and they currently aren't reassigned to anyone. For whatever reason it's messing up the query counts. I'd actually noticed the discrepancy before but haven't had time to dig into it. Now that I know whassup [sic] I might be able to fix that and get it accurate again.[/QUOTE]

Fixed. Okay, milestone page updated to show the countdown to <58M, and for the nitpickers I included the handful in the <56M and <57M ranges.

For purposes of showing how many are left, I'm going with the notion that even if there's a test for an exponent like [URL="http://www.mersenne.org/M57841879"]M57841879[/URL] where the one and only run is marked as 'suspect' in the database, then it doesn't really count? I haven't seen the assignment logic but in a case like that it may treat the exponent as a first-time check so it gets reassigned thusly (not waiting for the double-checkers).

Once all the #'s in a milestone are actually assigned it doesn't matter, but it threw it off a wee bit, just enough in this case to make it interesting.

Final 3 under 56M
 

Well, from what I can tell based on the progress of these 3 assignments, they all [I]should[/I] finish in time based on their rate of progress:
[CODE]54357769 2015-06-09 21:45
55027163 2015-06-30 13:48
55079077 2015-07-05 13:16[/CODE]

I'm keeping an eye on that 54M one since it's been pretty choppy, but dadgumit it's 99.2% done, yet it's going to expire pretty soon since it's 637 days old. The other 2 are a "mere" 556 days old and are at 85-86% done, and moving at a blazingly slow 0.33% per day or so. They're barely able to keep up with the exception for a grandfathered assignment in progress, but just barely.

Right now, at least, I expect all 3 should complete before they expire.

To remind, you already DC/TC those three, and they are not prime. But you did not report the result, waiting for the original assignee to report first and take the LL credit, then you will report and take the DC credit. This reminder is to avoid someone else wasting his resources on poaching them.

[QUOTE=LaurV;402724]To remind, you already DC/TC those three, and they are not prime. But you did not report the result, waiting for the original assignee to report first and take the LL credit, then you will report and take the DC credit. This reminder is to avoid someone else wasting his resources on poaching them.[/QUOTE]

Actually I've done my own test of that 54M exponent, but not the two 55M exponents. I have a better feeling about those 2 finishing so I wasn't too concerned (impatient of course, but not concerned that they'd expire first) :smile:

[QUOTE=Madpoo;402726]Actually I've done my own test of that 54M exponent, but not the two 55M exponents. I have a better feeling about those 2 finishing so I wasn't too concerned (impatient of course, but not concerned that they'd expire first) :smile:[/QUOTE]

FYI, those 2 exponents in the 54M range that I checked in (54674791, 54759797)... they're still not done. They would have expired a few days ago, but now at least the original user can keep on working on them and maybe they'll manage to finish them in another 4-5 years as a double-check, before they get assigned to someone else for that purpose. :smile:

Exponent M54357769 is still being sporadic. No check-ins since a few days ago (May 25) so it's still at 99.4%. It'll expire in about 15-16 days. I'm still hopeful this person will sneak in just under the wire but it's definitely a squeaker already. Kind of amazing that it's taken this person 650 days so far to get up to 99.4%, and may or may not finish before it expires.

Worst case, I'll check my result in just before expiration, and if they miss it by just a day or two, theirs will be a double-check.

[QUOTE=Madpoo;403217]Exponent M54357769 is still being sporadic. No check-ins since a few days ago (May 25) so it's still at 99.4%. It'll expire in about 15-16 days. I'm still hopeful this person will sneak in just under the wire but it's definitely a squeaker already. Kind of amazing that it's taken this person 650 days so far to get up to 99.4%, and may or may not finish before it expires.[/QUOTE]

Now up to 99.8%. Just two "ticks" to go!

Countdown to testing all exponents below M(57885161) once: 459
Countdown to first time checking all exponents below 58M: 459 (Estimated completion : 2015-10-22)

This suggests to me there are none between 57885161 and 58M?

[QUOTE=petrw1;403494]Countdown to testing all exponents below M(57885161) once: 459
Countdown to first time checking all exponents below 58M: 459 (Estimated completion : 2015-10-22)

This suggests to me there are none between 57885161 and 58M?[/QUOTE]

I believe it would suggest that all potential candidates between the two of which there are 6378 at very most have been checked at least once. edit: when you consider p%4==3 and isprime(2*p+1), 229 get eliminated with a non trivial factor for example.

[QUOTE=petrw1;403494]Countdown to testing all exponents below M(57885161) once: 459
Countdown to first time checking all exponents below 58M: 459 (Estimated completion : 2015-10-22)

This suggests to me there are none between 57885161 and 58M?[/QUOTE]

That's correct, all exponents between 57,885,161 and 58M have already been checked at least once. (EDIT: or factored, as suggested)

[QUOTE=cuBerBruce;403429]Now up to 99.8%. Just two "ticks" to go![/QUOTE]

I still have my fingers crossed. Weird to think this person will take nearly 660 days to finish this one test.

Here are my predictions on these final 4 under 57M:
[CODE]exponent Real Eta
54357769 2015-06-09 00:54
55027163 2015-07-18 07:53
55079077 2015-07-22 21:08
56931947 2015-06-13 13:57[/CODE]

I'm starting to have doubts about those two in the 55M range. They seem to have slowed down lately, and although I'm projecting 44 and 49 days left for them, they're at risk of expiring soon... they're 572 days old right now and as an example, if they stopped progressing entirely (unlikely), they'd expire in 53 and 50 days respectively. So even if they do progress a bit, they'd be struggling to keep up with the looming expiration deadline. They're averaging under 0.3 % per day (0.28% and 0.27%). They'd need to do a little better just to keep ahead of the game, like 0.3% per day.

That 56M exponent is a recent re-assignment when the original expired, so it'll have no problem finishing soon.

Have you tested any of these 4 numbers?
 

Have you tested any of these 4 numbers?

[QUOTE=Madpoo;403527]I still have my fingers crossed. Weird to think this person will take nearly 660 days to finish this one test.




Here are my predictions on these final 4 under 57M:
[CODE]exponent Real Eta
54357769 2015-06-09 00:54
55027163 2015-07-18 07:53
55079077 2015-07-22 21:08
56931947 2015-06-13 13:57[/CODE]

I'm starting to have doubts about those two in the 55M range. They seem to have slowed down lately, and although I'm projecting 44 and 49 days left for them, they're at risk of expiring soon... they're 572 days old right now and as an example, if they stopped progressing entirely (unlikely), they'd expire in 53 and 50 days respectively. So even if they do progress a bit, they'd be struggling to keep up with the looming expiration deadline. They're averaging under 0.3 % per day (0.28% and 0.27%). They'd need to do a little better just to keep ahead of the game, like 0.3% per day.

That 56M exponent is a recent re-assignment when the original expired, so it'll have no problem finishing soon.[/QUOTE]

M54357769 was finally finished by the person assigned to it after 654.8 days.

It's time for Madpoo to report his double-check of it.

[QUOTE=rudy235;403573]Have you tested any of these 4 numbers?[/QUOTE]

Madpoo has previously said he has done M54357769 without officially reporting the result, but I believe he denied working on the two in the 55M range. I think these two 55M have a decent chance of being finished by the assignee before expiring. Of course, there is every indication the fourth one will be finished soon by the current assignee.

The milestones report page is now out of date also.

1
 

So now all exponents bellow (or equal) to 55'000,000 have been tested at least once. :big grin:

[QUOTE=cuBerBruce;403607]M54357769 was finally finished by the person assigned to it after 654.8 days.

It's time for Madpoo to report his double-check of it.

Madpoo has previously said he has done M54357769 without officially reporting the result, but I believe he denied working on the two in the 55M range. I think these two 55M have a decent chance of being finished by the assignee before expiring. Of course, there is every indication the fourth one will be finished soon by the current assignee.[/QUOTE]

Yup, I just got online and saw that one *finally* finished. Just checked in my result which thankfully matched. It would have been a shame if it didn't after all the time that user put into it. :smile:

I haven't tested those 55M exponents, but if it's getting closer to the wire on those and they're at risk of expiring anyway, I may run them and have the results standing by.

I also just updated the milestones page with the June 6 date for everything under 55M.

[QUOTE=Madpoo;403620]...I haven't tested those 55M exponents, but if it's getting closer to the wire on those and they're at risk of expiring anyway, I may run them and have the results standing by.[/QUOTE]

Well, I was getting increasingly unsure that M55079077 would finish in time. The other one by that same user might since it was a whole 1% ahead and assigned on the same date, so it might squeak through.

I ran my own LL test of M55079077 and I'll just wait to check it in until it seems the other will expire, or as a double-check if they do manage to finish in time.

(spoiler alert: it's not prime)

If the other by that user also seems to be falling behind I'll probably do the same for it as well.

[QUOTE=Madpoo;403732]Well, I was getting increasingly unsure that M55079077 would finish in time. The other one by that same user might since it was a whole 1% ahead and assigned on the same date, so it might squeak through.[/QUOTE]

The two 55Ms were generally progressing at a significantly faster rate between April 23 and May 7 than either the weeks before or after that period. If these exponents get another period of similar progress in the next three months, these should get completed. But at the rate of progress since May 7, I project the one that's farther advanced to complete with only 5 days to spare.

All 56M range exponents now have had at least one LL (or factored). The two in the 55M range are still (at least officially) remaining. M56931947 has completed.

[QUOTE=cuBerBruce;403793]All 56M range exponents now have had at least one LL (or factored). The two in the 55M range are still (at least officially) remaining. M56931947 has completed.[/QUOTE]

Cool. Meaning we'll have another of those blips in the milestones where <56M and <57M are done on the same day. Oh well... it's happened before. It hopefully won't happen again since it was really just the grandfathered assignments holding things up, and there's precious few of those left.

There are 2 in the 58M range and 54 of them in the 59M range. I hope those expire or finish up soon. We'll be counting down to the <59M exponents before too long.

So, keep an eye on 58052513 and 58600229.

On the double-checking side, there aren't any in the 34M-37M range, I don't think. There's a couple in the 37M-38M range, but by the time we start counting those down, I bet they'll have expired.

[QUOTE=Madpoo;403810]Cool. Meaning we'll have another of those blips in the milestones where <56M and <57M are done on the same day. Oh well... it's happened before. [/QUOTE]
tl;dr The new system is working.

Recently:

Countdown to testing all exponents below M(57885161) once: [COLOR=blue]116
[/COLOR]Countdown to first time checking all exponents below 58M: [URL="http://www.mersenne.org/assignments/?exp_lo=55000000&exp_hi=58000000&execm=1&exp1=1&extf=1&exdchk=1"]116[/URL] [COLOR=green](Estimated completion : 2015-09-22)

If I remember correctly, there are 6378 primes between 57885161 and 58000000.

Does the fact that 116 = 116 mean that all of those primes have already been checked?

So that both milestones will fall simultaneously?

Clever people these GIMPsters.


[/COLOR]

[QUOTE=davar55;404898]Does the fact that 116 = 116 mean that all of those primes have already been checked?[/QUOTE]See post 1886 in this very thread.

Amazing what you can find these days.

remaining 55M exponents
 

[URL="http://www.mersenne.org/M55027163"]M55027163[/URL] and [URL="http://www.mersenne.org/M55079077"]M55079077[/URL] are increasingly unlikely to finish before expiring.

The progress on those has slowed greatly and right now it looks like they may finish up in 2 more months (63 and 66 days), but they'll expire before then (at the current rate, maybe in 35-45 days).

I've run my own tests on both of them and as before I'll hold those back until the 2 assignments for them are just about to expire.

[QUOTE=Madpoo;405479][URL="http://www.mersenne.org/M55027163"]M55027163[/URL] and [URL="http://www.mersenne.org/M55079077"]M55079077[/URL] are increasingly unlikely to finish before expiring.

The progress on those has slowed greatly and right now it looks like they may finish up in 2 more months (63 and 66 days), but they'll expire before then (at the current rate, maybe in 35-45 days).

I've run my own tests on both of them and as before I'll hold those back until the 2 assignments for them are just about to expire.[/QUOTE]And now, because you didn't report your result, the work has been wasted. [url=https://en.wikipedia.org/wiki/Mx_%28title%29]Mx[/url] ANONYMOUS has completed the exponent.
[url]http://www.mersenne.org/report_exponent/?exp_lo=55027163&full=1[/url]

Folks, whenever you do work, report it. Then there is lass wasted effort.

[QUOTE=retina;406084]And now, because you didn't report your result, the work has been wasted. [url=https://en.wikipedia.org/wiki/Mx_%28title%29]Mx[/url] ANONYMOUS has completed the exponent.
[url]http://www.mersenne.org/report_exponent/?exp_lo=55027163&full=1[/url]

Folks, whenever you do work, report it. Then there is lass wasted effort.[/QUOTE]

Do I still get to claim the moral high ground? I didn't poach it after all, by withholding my result. :smile:

In fact, just earlier today I did some projections on those 2 based on their rate of progress, where they are now, and the sliding "grace period" on them. Turns out they probably would have finished up in time, with a few days to spare even.

If it weren't for that I probably would have gone ahead and checked mine in. Seems like Mr. Anonymous didn't care, and didn't even have the guts to check it in as themselves probably.

I could figure out who the person is by looking at some logs... but I won't do it. I can tell that this person is probably going to check in a result for the other remaining 55M exponent, but I'd encourage them not to. I just saw some activity from that user looking into that one as well.

Let me just say that the original assignee, by my estimations, will finish those 2 exponents on August 31 and September 4. So just be patient, eh?

EDIT: I know for a fact this person is reading this thread, so... just hold off on doing that 2nd one. I've done my own pre-doublecheck and the original assignee will finish theirs ~ Sep 4

[QUOTE=Madpoo;406089]Do I still get to claim the moral high ground? I didn't poach it after all, by withholding my result. :smile:

In fact, just earlier today I did some projections on those 2 based on their rate of progress, where they are now, and the sliding "grace period" on them. Turns out they probably would have finished up in time, with a few days to spare even.

If it weren't for that I probably would have gone ahead and checked mine in. Seems like Mr. Anonymous didn't care, and didn't even have the guts to check it in as themselves probably.

I could figure out who the person is by looking at some logs... but I won't do it. I can tell that this person is probably going to check in a result for the other remaining 55M exponent, but I'd encourage them not to. I just saw some activity from that user looking into that one as well.

Let me just say that the original assignee, by my estimations, will finish those 2 exponents on August 31 and September 4. So just be patient, eh?

EDIT: I know for a fact this person is reading this thread, so... just hold off on doing that 2nd one. I've done my own pre-doublecheck and the original assignee will finish theirs ~ Sep 4[/QUOTE]
You probably know more about each contributor to GIMPS than anyone else on the planet, but ultimately no-one except the people themselves knows what motivates them and where they are "at". And, of course, they are indeed people, regardless of whether or not they identify themselves.:smile:

Yes, you have the moral high ground for not reporting your result before the original assignee finished. There is plenty of work for everyone, and with the expiry and reassignment policies in place there is really no excuse for anyone to poach other people's assignments.:tu:

[QUOTE=Brian-E;406105]You probably know more about each contributor to GIMPS than anyone else on the planet, but ultimately no-one except the people themselves knows what motivates them and where they are "at". And, of course, they are indeed people, regardless of whether or not they identify themselves.:smile:

Yes, you have the moral high ground for not reporting your result before the original assignee finished. There is plenty of work for everyone, and with the expiry and reassignment policies in place there is really no excuse for anyone to poach other people's assignments.:tu:[/QUOTE]

LOL... thanks. But I was kind of kidding because I've poached active assignments myself in the past, so I really can't claim to be all holier than thou. :smile: When I've done it before, I tried to be judicious and pick ones that seemed abandoned, but I've been wrong before.

That's why I setup a little thing to keep track of *actual* progress of these grandfathered things and try to guesstimate if it would expire before being done or not. There were just these last 3 (2 in the 55M and one in the 58M range) that were iffy and I thought would probably make it, so I was happy to let it play out.

Anyway, I wasn't really going to name and shame anyone since that's a little too "pot calling the kettle black" for me, I just wanted to make sure I encouraged the person to hold off on doing that other one if they had similar intentions.

[QUOTE=Madpoo;406124]Anyway, I wasn't really going to name and shame anyone since that's a little too "pot calling the kettle black" for me, I just wanted to make sure I encouraged the person to hold off on doing that other one if they had similar intentions.[/QUOTE]
Right! You're way ahead of everyone.:smile:

[QUOTE=Madpoo;406089]EDIT: I know for a fact this person is reading this thread, so... just hold off on doing that 2nd one. I've done my own pre-doublecheck and the original assignee will finish theirs ~ Sep 4[/QUOTE]I think the best way to discourage the rogue "perpetrator" from doing another wasted test is to report your result. Then there is no incentive to [dup|trip]licate the work.

[QUOTE=retina;406162]I think the best way to discourage the rogue "perpetrator" from doing another wasted test is to report your result. Then there is no incentive to [dup|trip]licate the work.[/QUOTE]

I can send you my result and let you have the honors. :smile: That way if anyone gets wee-wee'd up about it they can take it out on mean Mr. Retina. :devil:

[QUOTE=Madpoo;406180]I can send you my result and let you have the honors. :smile: That way if anyone gets wee-wee'd up about it they can take it out on mean Mr. Retina. :devil:[/QUOTE]Okay, no prob.. I can take the heat. Post your residue here. As long as you tell everyone else not to use it or read it then I can't see any difficulties. :unsure:

Seems like that user has stalled on their work. Nothing reported in for 2 weeks now on that last <57M assignment, and same on their (now) double check that got poached. Nothing heard since July 26.

At this rate, the first time check will expire in 8 more days based on the current % done. The double-check would expire in 11 days, although it wouldn't really expire then since double checks in the 55M range are a long ways out, so it would get some grace on that.

I'm thinking it might be time to call this and check in my result now. If the original user does check in again, I didn't have great confidence it was going to finish in time anyway... if it did it was going to be a squeaker for sure. Sep 7 was my last best estimate which put it very close to the max 665 days for the assignment age (about a day past). I'd be tempted to wait until the day before it expires just to give them every possible chance but I'll actually be away from my computer then.

So... here goes. I'm going to check in my double-check for that other one while I'm at it.

[QUOTE=Madpoo;407535]Seems like that user has stalled on their work. [/QUOTE]

Yup... that user still hadn't checked in any progress on [URL="http://www.mersenne.org/M55079077"]M55079077[/URL]

It would have expired on Aug. 16th (unless they actually do check in again and update the % done, but it probably still wouldn't have completed in time).

It would be awesome if the linked page showed the percentage completed and last check in of assigned work.

[QUOTE=Mark Rose;407988]It would be awesome if the linked page showed the percentage completed and last check in of assigned work.[/QUOTE]

Yeah, I kind of thought it'd be nice to have a shortcut to get to the assignment status of any given exponent, kind of like the /Mxxxx URL shortcut we setup for the detail of an exponent.

Right now it's "as easy as" going to a URL like:
[URL="http://www.mersenne.org/assignments/?exp_lo=55079077"]http://www.mersenne.org/assignments/?exp_lo=55079077[/URL]

Maybe a shortcut like /A55079077 that redirects to that would be handy. :smile:

[QUOTE]Countdown to testing all exponents below M(57885161) once: 8

Countdown to first time checking all exponents below 58M: 8 (Estimated completion : 2015-09-23)[/QUOTE]

Only [b][size=+3]8[/size][/b] to go!

While the top 2 of the 8 haven't checked in since mid-July, I would hope people will lay off from poaching any of these or working on them without being assigned to them. The existing assignments are actual cat 1 assignments, and should be completed or recycled soon enough.

I am curious that there always seems to be at least one exponent < 59M that remains unassigned. I would have to guess there is a single exponent that is in some sort of funny state that is preventing it from getting handed out. I don't know of any way to efficiently figure out which exponent that is from the public interfaces. I am wondering if anyone knows which exponent is unassigned, and why it remains unassigned.

[QUOTE=cuBerBruce;409122]I am curious that there always seems to be at least one exponent < 59M that remains unassigned. I would have to guess there is a single exponent that is in some sort of funny state that is preventing it from getting handed out. I don't know of any way to efficiently figure out which exponent that is from the public interfaces. I am wondering if anyone knows which exponent is unassigned, and why it remains unassigned.[/QUOTE]

I was curious about that too the other day, so I dug into it.

There's an exponent in the 58M range that, for some reason, is marked in the database as needing a double-check even though no first time check exists. I think it got tagged wrong for some old reason, like maybe a result did come in for it but was bogus or something, but meanwhile it never got set back to an exponent needing a first time check.

I'm waiting on George for confirmation or approval before setting it back to the "first time checks needed" category, just in case there was something else bizarre going on. I'm guessing just a weird data error. As soon as it's fixed it should get assigned out pretty quick. For now I didn't think there was any big hurry so there's no real rush to do anything before George weighs in. He does sometimes get results emailed to him and checks those in manually, and for all I know something like that happened and it just hasn't been entered in yet. :smile:

[QUOTE=Madpoo;409129]There's an exponent in the 58M range that, for some reason, is marked in the database as needing a double-check even though no first time check exists. I think it got tagged wrong for some old reason, like maybe a result did come in for it but was bogus or something, but meanwhile it never got set back to an exponent needing a first time check.[/QUOTE]
I can think of a rather exciting reason why an LL-test result might get hidden from general view until some rather urgent double, triple and quadruple tests had been carried out. Only theoretically, of course - I doubt that this is really the case here. Still, I do wonder what exactly [U]you[/U] would see in the database the next time that exciting circumstance were to happen (I believe that last time it happened you didn't have the intimate access to the database which you do now).

[QUOTE=Brian-E;409165]I can think of a rather exciting reason why an LL-test result might get hidden from general view until some rather urgent double, triple and quadruple tests had been carried out. Only theoretically, of course - I doubt that this is really the case here. Still, I do wonder what exactly [U]you[/U] would see in the database the next time that exciting circumstance were to happen (I believe that last time it happened you didn't have the intimate access to the database which you do now).[/QUOTE]

Well, it was nothing as exciting as that. The exponent in question is: M58138603

Neither George nor I had any explanation for why it was marked as "double check needed" instead of a first time check, but anyway, it got fixed. If it hasn't been already, it'll be assigned to someone pretty soon.

As for what happens when a prime is discovered, I'm not really sure what all takes place... I know there are certain things that happen to try and keep word of that from leaking out until confirmations are done, and I've added a few things myself to just make sure none of the reports will accidentally leak that info prematurely.

I think a determined person could probably figure it out if they searched hard enough, but I'm not really sure... :smile:

[QUOTE=Madpoo;409212]The exponent in question is: M58138603[/QUOTE]

If's the lowest of the remaining 58Ms, and yes, it's been assigned to someone. It has bumped my #13 exponent down to #14.

[QUOTE=Madpoo;409212]As for what happens when a prime is discovered, I'm not really sure what all takes place... [/QUOTE]
That is very easy to find, we only have to find a new prime. So, everybody back to work, and turn that prime in, till tonight! :smile:

[QUOTE=LaurV;409233]That is very easy to find, we only have to find a new prime. So, everybody back to work, and turn that prime in, till tonight! :smile:[/QUOTE]

Maybe. :smile: I remember looking at past conversations when people suspected a prime was found. Let's just say that some methods used previously to divine the exponent by looking for what wasn't there may not be so successful. There still might be other ways though, I haven't really gone through it all.

[QUOTE] Countdown to testing all exponents below M(57885161) once: 7

Countdown to first time checking all exponents below 58M: 7 (Estimated completion : 2015-09-22)
[/QUOTE]

[size=+3][b]7[/b][/size] to go...

It looks to me that M(57404873) was not really poached, but rather completed by someone who was assigned it once, but didn't complete it within the required 90 days.

[offtopic]
grrr... I looked 5 minutes to your avatar, even enlarged it till it was matching my pupillary distance and looked more, with my eyes crossed... It is not a 3D animation, for sure! :razz:
[/offtopic]

[QUOTE=LaurV;409422][offtopic]
grrr... I looked 5 minutes to your avatar, even enlarged it till it was matching my pupillary distance and looked more, with my eyes crossed... It is not a 3D animation, for sure! :razz:
[/offtopic][/QUOTE]
[noparse]
[offtopic]
For the record, I have not set any avatar. I generally find them too distracting and disable them, so if an admin messed around with my avatar, I probably wouldn't even notice.
[/offtopic]
[/noparse]

The number of exponents between 1.000.000 and 2.000.000 that have not been factored and hence 'only' have a LL test has dropped to the nice round value of 20.000 in the hourly stats page.

-

Maybe a milestone worth mentioning is the fact that by today, there is [URL="http://www.mersenne.ca/status/tf/0/2/0"]no exponent TFed under 62[/URL]. The last two of them [URL="http://www.mersenne.ca/status/tf/2/4/0"]moved from 61 to 62[/URL] yesterday (or the day before, depending on your time zone).

In this respect, I think that James' fresh tables would have to be edited, to get rid of the "<61" and "61" columns, making it more suitable to fit the screen for some of us who are still using narrow monitors (the upper right menu contributes to this too, pushing wider tables down the page, but that is another story). We "know" there are no expos under 62, so there should not be necessary to have a column for it. If the tickle in the palm is too unbearable, then other columns can be added on the other side (like 80++ bits, useful for LMH).

(sorry James :razz:)

Because things were working so well and others were watching the pot I delayed an update until today.

All exponents below [B][COLOR="Blue"]34,661,527[/COLOR][/B] have been tested and double-checked.
All exponents below [B][COLOR="Blue"]57,861,131[/COLOR][/B] have been tested at least once.

Countdown to testing all exponents below M([B][COLOR="Blue"]57885161[/COLOR][/B]) once: [COLOR="Red"][B]1[/B][/COLOR]
Countdown to first time checking all exponents below 58M: [B][COLOR="Red"]1[/COLOR][/B] (Estimated completion : [COLOR="Green"]2015-10-02[/COLOR])
Countdown to first time checking all exponents below 59M: [B][COLOR="Red"]47[/COLOR][/B] (Estimated completion : [COLOR="Green"]2015-11-02[/COLOR])
Countdown to double checking all exponents below 35M: [B][COLOR="Red"]3,421[/COLOR][/B] (2,868 still unassigned)

Countdown to proving M([COLOR="Green"]37156667[/COLOR]) is the [COLOR="green"]45[/COLOR]th Mersenne Prime: 27,643[/QUOTE]

[/QUOTE]?
:razz:

[QUOTE=Uncwilly;411683]Because things were working so well and others were watching the pot I delayed an update until today.

Countdown to first time checking all exponents below 58M: [B][COLOR="Red"]1[/COLOR][/B] (Estimated completion : [COLOR="Green"]2015-10-02[/COLOR])[/QUOTE]

If you were talking about the single remaining test under 58M...

I poached one of them. It hadn't reported in for about a month, hadn't started yet at the time. It seemed abandoned by user "Xebecer"? Well, that assignment is now a DC.

If my run had shown it was prime, I think I would have let George know but waited to see if the original assignee checked in so they could "discover" it (I don't need the prize money), but it was composite anyway.

[QUOTE=Madpoo;411694]If you were talking about the single remaining test under 58M...

I poached one of them. It hadn't reported in for about a month, hadn't started yet at the time. It seemed abandoned by user "Xebecer"? Well, that assignment is now a DC.[/QUOTE]Perhaps that is a bug in the recycling rules implementation code? It appeared to be a manual check-in because it had a finish date before the most recent check-in. But it somehow got to 70+ days old with no progress without triggering any recycling.

[QUOTE=Madpoo;411694]If you were talking about the single remaining test under 58M...

I poached one of them. It hadn't reported in for about a month, hadn't started yet at the time. It seemed abandoned by user "Xebecer"? Well, that assignment is now a DC.

If my run had shown it was prime, I think I would have let George know but waited to see if the original assignee checked in so they could "discover" it (I don't need the prize money), but it was composite anyway.[/QUOTE]

Oh, and FYI, my own analysis of M57861131 (the single remaining test) has an ETA of Oct. 5, not Oct. 2. That's based on an observed progress rate of 5.1% daily over the past 13 days.

[QUOTE=retina;411695]Perhaps that is a bug in the recycling rules implementation code? It appeared to be a manual check-in because it had a finish date before the most recent check-in. But it somehow got to 70+ days old with no progress without triggering any recycling.[/QUOTE]

Yeah, I think it had something to do with the client's "next check in date" that it sent at it's last communication. For some reason it was set to a time *after* it was due to be finished, which was weird. I guess you could configure a client to only communicate with Primenet once every few months, but it would finish work before then. Weird at any rate.

[QUOTE=Uncwilly;372515]Straight scale, with primes, and rich data.[/QUOTE][QUOTE=Uncwilly;372462]Well, I have fairly good data from ~Oct 2010. Is this good enough for the current period?[/QUOTE]Updates on the 2 graphs noted in the above posts:

Oh, you mean that is the time to find a new prime, arn't you? :wink:

Milestone page update (under the hood)
 

Hey all,

I've been working a little on the milestone page to make it show the stats a little faster.

Let me give some quick behind the scenes... Currently, every time that page is loaded it has to query some stuff from SQL for each of the countdown stats. It's not a huge burden, but it does add up. When we had a false positive the other day for a prime, it got me to thinking about what would happen if a new prime is found, and the site starts getting a bunch of traffic?

The home page, download page, history, etc. would probably get a surge in traffic and those are all pretty much static, but that milestone page would also probably get more views and that would probably be a bad thing.

Solution was to cache the stats instead of a query each time it's viewed. Right now I have it cache for 5 minutes at a time. George had mentioned some other possible things we could do with that, but for now if you want to check it out:
[URL="http://www.mersenne.org/report_milestones/default.mock.php"]http://www.mersenne.org/report_milestones/default.mock.php[/URL]

There's a little footer at the bottom of each page that shows the time it took the server to render the HTML. For the cached version it's a fraction of a second (assuming you got the cached info... you may be the lucky person who views it and triggers a fresh query). We're talking 60-70 milliseconds. The old version is taking me an average 2.4 - 2.5 seconds. Quite a difference.

Any feedback on this? Is 5 minutes fine? The milestones don't typically change that drastically and I only set it to 5 since the <58M milestone is approaching... if we didn't have any pending completions I might set it higher like an hour or two.

Enjoy! Once we've fleshed it out a bit we'll move that new scheme so it replaces the live page. The page itself is the same except for the text that shows the date/time the stats were last fetched, so that's how you'll know when it's live.

[QUOTE=Madpoo;411846]
Any feedback on this? Is 5 minutes fine? The milestones don't typically change that drastically and I only set it to 5 since the <58M milestone is approaching... if we didn't have any pending completions I might set it higher like an hour or two.[/QUOTE]

Once an hour is fine in my opinion

[QUOTE=petrw1;411849]Once an hour is fine in my opinion[/QUOTE]

In my experience with caches for this kind of thing, there's very little difference between five and sixty minutes. If the page is only hit a few times per minute, there's very little system impact even with no caching. So if you can cache for a minute, you're good. When the system gets busy, say 10 hits per second in this case, then the caching pays off by producing a handful of requests per five minutes instead of 3000 requests.

So why a handful of requests? The problem is the [url=https://en.wikipedia.org/wiki/Thundering_herd_problem]thundering herd problem[/url]. When the cache suddenly expires, [i]all[/i] the parallel requests will start rebuilding it. There are mitigation techniques, but if the system can handle the thundering herd until the cache is rebuilt, the mitigations are not always worth implementing. The most important thing is to make sure the cache is written atomically so two parallel generations don't interfere.

[QUOTE=petrw1;411849]Once an hour is fine in my opinion[/QUOTE]I concur. I believe that I previously expressed this opinion a while back. The 'classic' stats page is done once an hour, right? Doing all of the stats around 5 minutes before the hour (including the work distribution report) would be fine. Only serve up fresh stats on drill downs and specific requests.

Yaaaarrrrr :chappy:
Feeling good to be part of a milestone :razz:
[QUOTE]2014-08-09 First 100 million digit number successfully, and independently, double-checked - M(345,678,877).
[/QUOTE]

[ontopic]
One hour is ok.

1. Make it 15 minutes. Not too long, not too often. Just right!
2. Rebuild the cache on schedule, instead of off a page request.
EDIT:-
3. If needed, for the "critical" exponents, add custom code to rebuild the cache instantly when they're checked in

When was the last time?
 

If things go as expected and the mersenne number with exponent 57861131 is found composite and no new Mersenne Prime magically appears in the next 48 hours...

When was the last time that all mersenne exponents up to the largest mersenne prime –in this particular case M(48)– have been checked once and found composite?

I am guessing that that has not been the case for the last 50 years at least. (yes, I realize that checking once does not make 100% certain that there in not another prime "hidden" somewhere in there.)

Another trivial question.

In 1988 Colquitt & Welsh discovered M(29) exponent 110503 a 5 years after M(30) and 3 years after (M31).

Was this the case of a first LL test that gave composite and no one doing a second LL? Seems a bit odd that 5 years passed (1985-1988) without that exponent being checked again. :ermm: