Can I cut an ECM assignment short?

[Mr. P-1]

I took this assignment from the server, out of curiosity as to how this type of work proceeds:

ECM2=6B2185CA275DF9CFED1B2B434D64A7AF,1,2,2233513,-1,50000,5000000,150

This admittedly rather old thread confirms my suspicion that the last parameter, 150, is the number of curves to do.

I don't wish to do that many, but I do want the dozen or so I have done to be recorded. Can I safely edit worktodo.txt to do fewer curves without upsetting primenet?

[MatWur-S530113]

Hello,

I never lowered the number of curves. I would unreserve the exponent, contact the server for it, and then exit prime95,
add the line
ECM2=1,2,2233513,-1,50000,5000000,xx
to the worktodo.txt and run prime95 again, contact the server again and you will get a new assignment. (Of course reset 'xx' with the number of curves you want to run).

Best regards,

Matthias

[Mr. P-1]

Cheers. I'll do that.

[henryzz]

i expect you can
today i did some experimentation with tf assignments and i reduced the limit i was factoring to and it worked fine
if you look on http://www.mersenne.org/workload/ then you will see that there is nothing that says how many curves you are doing or what you are tfing to

[MatWur-S530113]

Hello,

I already did it with increasing the number of curves. That worked fine, too. The correct number of curves were reportet, the correct CPU-credit calculatet and the assignment was correct deletet after reporting the result. But as I said, I never decreased the number of curves before. If you had no problems with decreasing the tf-limit I think this will work, too.
As you said, the assignments only show the kind of works, but no tf-limit, boundary or number of curves. I suspect that this information is not stored in the AID.

best regards,

Matthias

[Mr. P-1]

I just reduced the number of curves without unreserving. The result was credited on PrimeNet just fine.

[ckdo]

If anyone was wondering what would happen if you reduced the number of curves below the count that has been done already... see pic.

[cheesehead]

Originally Posted by ckdo

If anyone was wondering what would happen if you reduced the number of curves below the count that has been done already... see pic.

So, (just to be sure of interpretation) the estimated time-to-completion for 2450993 was a negative 67 minutes (08:48 minus 09:55)?

- - -

(Interestingly, the first clue my employer had about the Pentium FDIV bug, shortly before public announcements reached us, was a report showing a negative elapsed time for processing an ATM transaction.)