GPU to 72 assignments completed prematurely?

[ixfd64]

I requested 30 TF assignments via GPU to 72 about a week ago. I've completed 25 of them so far, and I plan to submit them all at once. But when I checked my assignments page, it shows all of them as completed.

What gives?

[petrw1]

Originally Posted by ixfd64

I requested 30 TF assignments via GPU to 72 about a week ago. I've completed 25 of them so far, and I plan to submit them all at once. But when I checked my assignments page, it shows all of them as completed.

What gives?

All by the same person???

[chalsall]

Originally Posted by ixfd64

I requested 30 TF assignments via GPU to 72 about a week ago. I've completed 25 of them so far, and I plan to submit them all at once. But when I checked my assignments page, it shows all of them as completed.

What gives?

Could you give a couple of examples?

Records are kept for all assignments, so if you let me know which assignments you're talking about, I can found out what happened.

[ixfd64]

Well, I just found out what happened. It seems they were poached by "Kevin Jaget": http://www.mersenne.org/report_expon...xp_lo=59752981

[flashjh]

Originally Posted by chalsall

Could you give a couple of examples?

Records are kept for all assignments, so if you let me know which assignments you're talking about, I can found out what happened.

More expiration overlap?

[chalsall]

Originally Posted by flashjh

More expiration overlap?

Yup, I'm afraid so.

Thanks to Garo's pointing out that the expiry policy is no longer really needed, this won't happen any more.

[ixfd64]

From my experience, PrimeNet will reject TF results if the range has already been tested. However, I've seen a couple of exponents that have overlapping TF results.

Is there some sort of trick to force PrimeNet to accept these results, or do I have to ask George to manually enter them into the database?

[ixfd64]

Sure enough, PrimeNet is rejecting my results. Time to shoot George a PM.

[ixfd64]

Good news and bad news. I was able to trick PrimeNet into accepting my results by changing the factor depth to 74 bits in my results.txt file. Unfortunately, the results are partially incorrect as I didn't really TF to 74 (or even 73) bits.

George or Scott, if you happen to see this post, please change the "2^74" to "2^72" and adjust my GHz-days accordingly. Thanks!

The correct results should be:

[Mon Nov 26 22:53:22 2012]
UID: ixfd64/andromeda, no factor for M59752981 from 2^71 to 2^72 [mfaktc 0.19 barrett76_mul32]
[Tue Nov 27 21:08:34 2012]
UID: ixfd64/andromeda, no factor for M59753069 from 2^71 to 2^72 [mfaktc 0.19 barrett76_mul32]
[Wed Nov 28 18:47:40 2012]
UID: ixfd64/andromeda, no factor for M59753131 from 2^71 to 2^72 [mfaktc 0.19 barrett76_mul32]
[Wed Nov 28 21:45:23 2012]
UID: ixfd64/andromeda, no factor for M59753179 from 2^71 to 2^72 [mfaktc 0.19 barrett76_mul32]
[Thu Nov 29 19:42:14 2012]
UID: ixfd64/andromeda, no factor for M59753209 from 2^71 to 2^72 [mfaktc 0.19 barrett76_mul32]
[Thu Nov 29 22:57:48 2012]
UID: ixfd64/andromeda, no factor for M59753219 from 2^71 to 2^72 [mfaktc 0.19 barrett76_mul32]
[Fri Nov 30 19:37:17 2012]
UID: ixfd64/andromeda, no factor for M59753279 from 2^71 to 2^72 [mfaktc 0.19 barrett76_mul32]
[Fri Nov 30 22:32:17 2012]
UID: ixfd64/andromeda, no factor for M59753329 from 2^71 to 2^72 [mfaktc 0.19 barrett76_mul32]
[Sat Dec 01 01:23:12 2012]
UID: ixfd64/andromeda, no factor for M59753339 from 2^71 to 2^72 [mfaktc 0.19 barrett76_mul32]
[Sat Dec 01 04:14:24 2012]
UID: ixfd64/andromeda, no factor for M59753341 from 2^71 to 2^72 [mfaktc 0.19 barrett76_mul32]
[Sat Dec 01 07:05:35 2012]
UID: ixfd64/andromeda, no factor for M59753357 from 2^71 to 2^72 [mfaktc 0.19 barrett76_mul32]
[Sat Dec 01 09:56:48 2012]
UID: ixfd64/andromeda, no factor for M59753377 from 2^71 to 2^72 [mfaktc 0.19 barrett76_mul32]
[Sat Dec 01 12:47:57 2012]
UID: ixfd64/andromeda, no factor for M59753479 from 2^71 to 2^72 [mfaktc 0.19 barrett76_mul32]
[Sat Dec 01 15:39:07 2012]
UID: ixfd64/andromeda, no factor for M59753539 from 2^71 to 2^72 [mfaktc 0.19 barrett76_mul32]
[Sat Dec 01 18:30:21 2012]
UID: ixfd64/andromeda, no factor for M59753663 from 2^71 to 2^72 [mfaktc 0.19 barrett76_mul32]
[Sat Dec 01 21:21:36 2012]
UID: ixfd64/andromeda, no factor for M59753693 from 2^71 to 2^72 [mfaktc 0.19 barrett76_mul32]
[Sun Dec 02 00:12:43 2012]
UID: ixfd64/andromeda, no factor for M59753741 from 2^71 to 2^72 [mfaktc 0.19 barrett76_mul32]
[Sun Dec 02 03:03:42 2012]
UID: ixfd64/andromeda, no factor for M59756927 from 2^71 to 2^72 [mfaktc 0.19 barrett76_mul32]
[Sun Dec 02 05:54:33 2012]
UID: ixfd64/andromeda, no factor for M59756951 from 2^71 to 2^72 [mfaktc 0.19 barrett76_mul32]
[Sun Dec 02 08:45:25 2012]
UID: ixfd64/andromeda, no factor for M59757041 from 2^71 to 2^72 [mfaktc 0.19 barrett76_mul32]
[Sun Dec 02 11:36:13 2012]
UID: ixfd64/andromeda, no factor for M59757149 from 2^71 to 2^72 [mfaktc 0.19 barrett76_mul32]
[Sun Dec 02 14:27:08 2012]
UID: ixfd64/andromeda, no factor for M59757209 from 2^71 to 2^72 [mfaktc 0.19 barrett76_mul32]
[Sun Dec 02 17:18:04 2012]
UID: ixfd64/andromeda, no factor for M59757319 from 2^71 to 2^72 [mfaktc 0.19 barrett76_mul32]
[Sun Dec 02 20:09:00 2012]
UID: ixfd64/andromeda, no factor for M59757391 from 2^71 to 2^72 [mfaktc 0.19 barrett76_mul32]
[Sun Dec 02 22:59:55 2012]
UID: ixfd64/andromeda, no factor for M59757433 from 2^71 to 2^72 [mfaktc 0.19 barrett76_mul32]
[Mon Dec 03 19:26:26 2012]
UID: ixfd64/andromeda, no factor for M59757449 from 2^71 to 2^72 [mfaktc 0.19 barrett76_mul32]
[Mon Dec 03 21:08:49 2012]
UID: ixfd64/andromeda, no factor for M59757571 from 2^71 to 2^72 [mfaktc 0.19 barrett76_mul32]
[Mon Dec 03 22:49:28 2012]
UID: ixfd64/andromeda, no factor for M59757587 from 2^71 to 2^72 [mfaktc 0.19 barrett76_mul32]
[Tue Dec 04 19:16:03 2012]
UID: ixfd64/andromeda, no factor for M59757713 from 2^71 to 2^72 [mfaktc 0.19 barrett76_mul32]
[Tue Dec 04 20:57:39 2012]
UID: ixfd64/andromeda, no factor for M59757721 from 2^71 to 2^72 [mfaktc 0.19 barrett76_mul32]

Edit: I suppose I could make these results valid by TF'ing the [73, 74] range for real. Unfortunately, the GTX 555 isn't the fastest GPU in the world, and it'll take me a long time to do so. Volunteers welcome, of course. :-)

[flashjh]

Originally Posted by ixfd64

I was able to trick PrimeNet into accepting my results by changing the factor depth to 74 bits in my results.txt file.

[garo]

That is an absolute no no! You shopuld have mailed your relevant results to George and he would have processed them manually. Now we have crappy data on the server.