(archive)MISFIT - Page 12

swl551 · 2012-12-25, 20:49

Quote:

Originally Posted by kracker

In the leader board, is it possible to display specific categories? ex. (rank in TF, P-1, LL etc)

VERSION 2.2.7
1. Leaderboard now allows selection of GIMPS_OVERALL and GIMPS_TF statistics. I am only adding TF since it is the only statistic, other than ALL, that is relevant to MISFIT.

Get from http://www.mersenneforum.org/misfit

kracker · 2012-12-25, 21:33

Quote:

Originally Posted by swl551

VERSION 2.2.7
1. Leaderboard now allows selection of GIMPS_OVERALL and GIMPS_TF statistics. I am only adding TF since it is the only statistic, other than ALL, that is relevant to MISFIT.

Get from http://www.mersenneforum.org/misfit

Wha.. Noooo... Why did you select my pathetic position??

...

Thanks, as always looking good.

swl551 · 2012-12-25, 21:38

Quote:

Originally Posted by kracker

Wha.. Noooo... Why did you select my pathetic position??

...

Thanks, as always looking good.

If I picked FLASHJH everyone would feel bad....

kracker · 2012-12-25, 23:17

Quote:

Originally Posted by swl551

If I picked FLASHJH everyone would feel bad....

Everyone knows he has a whole house full of gpu's crammed he doesn't live there! So, not everyone1!

Except nucleon. Maybe.

Dubslow · 2012-12-26, 02:11

Quote:

Originally Posted by kracker

Except nucleon. Maybe.

Yes, except he's shut down for the summer. Wait till ~March...

flashjh · 2012-12-26, 17:08

@Scott: I figured out what's happening with the difference between G72 and MISFIT.

I have StopAfterFactor=0 so that each range gets fully tested. Since a typical instance takes about 6-8 hours to complete in this range, sometimes a factor is found and reported in the results.txt file but the range isn't done when the time comes for MISFIT to submit.

When the exponent is complete a line like the following is put in the results.txt file:

Code:

found 1 factor for M60530593 from 2^71 to 2^73 [mfaktc 0.19 barrett76_mul32]

MISFIT uses that line to calculate factors found and adds 1 (or whatever was found) to the completed even though the actual factor has already been submitted and G72 removed it from the assignments.

I suppose it would require quite a lot of changes to the results code and how MISFIT handles a factor if a range isn't complete. It does explain why I haven't been able to find the duplicate(s) over the last few days (which has varied from 0 to 2).

I don't think it's that big of a deal. Even if another factor is found in the same range after MISFIT submits, PrimeNet will still accept more factors (though I'm not sure if G72 will pickup the extra factor, that isn't a big deal either).

swl551 · 2012-12-26, 20:21

Quote:

Originally Posted by flashjh

@Scott: I figured out what's happening with the difference between G72 and MISFIT.

I have StopAfterFactor=0 so that each range gets fully tested. Since a typical instance takes about 6-8 hours to complete in this range, sometimes a factor is found and reported in the results.txt file but the range isn't done when the time comes for MISFIT to submit.......

Well that explains why the change to uploading once per day had an impact. This scenario was no where on my radar!

...
MISFIT would have to look into Results.Txt and make comparisons to exponents in WorkToDo.txt to know if it "OK" to export certain results. The increased overhead of this processing would add time to the export function. Until MFAKTC has file locking I think this is off the table.

However it may not be worth it anyway since it really isn't harming anything.

Very interesting though....

flashjh · 2012-12-26, 21:09

Quote:

Originally Posted by swl551

Very interesting though....

Agreed, and it only affects someone with StopAfterFactor=0 anyhow.

NBtarheel_33 · 2012-12-27, 15:20

Quote:

Originally Posted by swl551

Since the first pubic release to the latest, hundreds of hours of work has been spent improving things.

Calling Mr. Eddy, Mr. David Eddy...

kracker · 2012-12-27, 15:38

Quote:

Originally Posted by NBtarheel_33

Calling Mr. Eddy, Mr. David Eddy...

swl551 · 2012-12-27, 17:35

Quote:

Originally Posted by NBtarheel_33

Calling Mr. Eddy, Mr. David Eddy...

I put that in there on purpose to see if anyone was paying attention!

2012-12-26, 17:08	#127
flashjh "Jerry" Nov 2011 Vancouver, WA 1,123 Posts	@Scott: I figured out what's happening with the difference between G72 and MISFIT. I have StopAfterFactor=0 so that each range gets fully tested. Since a typical instance takes about 6-8 hours to complete in this range, sometimes a factor is found and reported in the results.txt file but the range isn't done when the time comes for MISFIT to submit. When the exponent is complete a line like the following is put in the results.txt file: Code: found 1 factor for M60530593 from 2^71 to 2^73 [mfaktc 0.19 barrett76_mul32] MISFIT uses that line to calculate factors found and adds 1 (or whatever was found) to the completed even though the actual factor has already been submitted and G72 removed it from the assignments. I suppose it would require quite a lot of changes to the results code and how MISFIT handles a factor if a range isn't complete. It does explain why I haven't been able to find the duplicate(s) over the last few days (which has varied from 0 to 2). I don't think it's that big of a deal. Even if another factor is found in the same range after MISFIT submits, PrimeNet will still accept more factors (though I'm not sure if G72 will pickup the extra factor, that isn't a big deal either).

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
questions about misfit	jschwar313	MISFIT	1	2016-02-02 00:32
Help with Misfit	TheMawn	MISFIT	18	2015-06-21 22:00
Misfit feature request.	Aramis Wyler	MISFIT	15	2015-03-04 21:13
MISFIT 2.9	swl551	MISFIT	44	2014-12-15 12:46
MISFIT for MFAKTx	swl551	MISFIT	186	2014-02-05 23:27