Newer milestone thread

[ATH]

I just manually looked at 200k or 300k ranges between 83M and 86,028,121 whatever showed below 1000 results for each page and added them up manually.

I suggested this milestone previously for the milestone page, but since it is not there, I just did it manually instead.


Edit: Nevermind, I'm an idiot, I counted on the assignment lists, I forgot all the unassigned exponents, so the real number is quite a bit higher.

Another idea for a milestone:

Countdown to verifying all tests below M(49 979 687) (3,000,000th prime exponent)

[rudy235]

Originally Posted by ATH

I just manually looked at 200k or 300k ranges between 83M and 86,028,121 whatever showed below 1000 results for each page and added them up manually.

I suggested this milestone previously for the milestone page, but since it is not there, I just did it manually instead.


Edit: Nevermind, I'm an idiot, I counted on the assignment lists, I forgot all the unassigned exponents, so the real number is quite a bit higher.

I did not want to question your amount but to me, it looked to be over 10,000

[rudy235]

Does anyone know what is happening with this exponent?

http://https://www.mersenne.org/report_exponent/?exp_lo=83486549&full=1

It's been in the books for over 15 months!
It says it's supposed to finish in 1 day more, but it does not say when it might expire. Of course, I have no idea if it is advancing but somehow it has managed -without getting there- to avoid expiring for 15 months.
Also, it is a PRP search so what happens is someone does an LL test? Will the PRP result (if it finishes) count as 1 of the 2 needed for verification?

[Uncwilly]

I suspect that someone will take note of the exponent and will do a PRP and wait to submit it. If it expires, it will be submitted as a first time. If it does not expire, it will be an early double.

[rudy235]

Originally Posted by Uncwilly

I suspect that someone will take note of the exponent and will do a PRP and wait to submit it. If it expires, it will be submitted as a first time. If it does not expire, it will be an early double.

Yes, that's par for the course but I wonder if it has been progressing lately or not.

[ATH]

Actually I'm about to "poach" it in ~8 hours, since it has no Expire Date and as you said it has been going for 15.5 months. If it had been an LL assignment on Jan 7th 2018 it would have been Cat 3, and would have expired after 270 days on Oct 4th 2018:
https://www.mersenne.org/thresholds/?dt=2018-01-07

I think all PRP assignment had no expiration limit until recently?

Edit: Cat 4 actually, but would still have expired after 360 days on January 2nd 2019.

[kriesel]

Originally Posted by rudy235

Does anyone know what is happening with this exponent?

http://https://www.mersenne.org/report_exponent/?exp_lo=83486549&full=1

It's been in the books for over 15 months!
It says it's supposed to finish in 1 day more, but it does not say when it might expire. Of course, I have no idea if it is advancing but somehow it has managed -without getting there- to avoid expiring for 15 months.
Also, it is a PRP search so what happens is someone does an LL test? Will the PRP result (if it finishes) count as 1 of the 2 needed for verification?

PRP expiration is a recent phenomenon. This one was issued well before expiration for PRP was implemented. The server keeps track of expiration on a per-exponent basis, and apparently expiration-proof PRP assignments on a per-exponent basis also. I wonder if this particular run has gotten stuck in a GEC fail/retry loop, or a different exponent gotten manually inserted ahead of it in the worktodo file.
If someone does an LL test, or a PRP test of a different residue type, there will need to be another run of some matching type to see whether there's a residue match. Could be LL, could be PRP same-residue-type.

Note, I have an example of primenet not handling well, the case of PRP DC. I guess it takes a while to work the kinks out of the code when adding something like PRP support to a running system.
See https://www.mersenne.org/report_expo...9335979&full=1 It has one each PRP type 1 and PRP type 4 completed already by March 27.

I got assigned March 28 that exponent on prime95 on an i3-370M (slow) system after setting it to request PRP DC from primenet, so it should have almost a year to complete it as a Cat 4 DC. Its expiration date shown in 5 days and being listed as issued as Cat 0 in https://www.mersenne.org/workload/ show it's being treated like a Cat 0 first time PRP, despite already having TWO completed PRPs before being assigned in response to PRPDC assignment preference! Prime95 does not show whether it's running type 1 PRP, type 4 PRP, or some other type. According to https://www.mersenne.org/thresholds/ it should be listed and treated as Cat 4 PRP DC. (The Double Check heading does not specify LL, PRP, or both. The first test section does say "Lucas Lehmer" not "primality first test" or "Lucas Lehmer or PRP".) Upgrading to prime95 v29.7b1 and then v29.8b1 has not resolved it.

[Madpoo]

Originally Posted by kriesel

...I got assigned March 28 that exponent on prime95 on an i3-370M (slow) system after setting it to request PRP DC from primenet, so it should have almost a year to complete it as a Cat 4 DC. Its expiration date shown in 5 days and being listed as issued as Cat 0...

I'm not sure how or why, but it was assigned as a first-time PRP check.

I manually switched it to a double-check and you should be good. I wonder if that had anything to do with the mismatched set of previous results on it. The bit of code that assigns work as either first or double-check may have run into something funny because of that.

[GP2]

Originally Posted by kriesel

Prime95 does not show whether it's running type 1 PRP, type 4 PRP, or some other type.

Look at the PRP= (or PRPDC=) line in your worktodo.txt file. Does it end in ,3,1 or ,3,4 ?

[kriesel]

Originally Posted by Madpoo

I'm not sure how or why, but it was assigned as a first-time PRP check.

I manually switched it to a double-check and you should be good. I wonder if that had anything to do with the mismatched set of previous results on it. The bit of code that assigns work as either first or double-check may have run into something funny because of that.

Thanks for the immediate fix. It looks like you either caught it in the nick of time, in the last day before expiration, or unexpired it during the fix. It's now showing up as a PRPDC in my assignments list, as PRP-D, cat 4, 272 days to go, and no expiration indicated. It's probably worthwhile for someone to investigate how the assignment code let that PRP status happen, since activity is shifting from LL toward PRP, and while prime95 seems to favor type 1, gpuowl of recent versions seem to favor type 4, so it's likely to come up more often as time goes on.

Originally Posted by GP2

Look at the PRP= (or PRPDC=) line in your worktodo.txt file. Does it end in ,3,1 or ,3,4 ?

PRP=, and ,3,1. But since prime95 is a GUI application on Windows, I'm of the opinion that it ought rarely be necessary to go spelunking in the underlying files to sleuth out what happened. A worker window title saying in effect, PRP or LL, first time or DC, residue type 1 or 4 or whatever, % complete, whatever doesn't specify the work in the window contents at least. Heck, even EDT on a DEC VT52 text terminal in 1981 would show some status info in a fixed location on screen.

[kriesel]

forrest-lumpkin please double check your double-check runs (34 stalled since March in the 47m-48M range)