Results not recorded?

[Yura]

I sent 2 finished primes some 10 days ago using prime95 after few gruelling months but they are not recorded on my account. Could somebody guide/explain me why is that happening and what I do wrong.

I did over the years manual testing and sending results to George by e-mail, than had a pause of a year or two and now I am back. I process primes on PC#1 (no Internet) 24/7 and after they are done, copy the whole folder and send result from home, from PC#2.

[lycorn]

Check the "Results queries -> Exponent Status" on Primenet web page for the exponents you sent. Let us know what their current status is.
Hint: If you submitted the results manually (via the Manual Pages) you had to be logged in to receive the credit. Otherwise it goes to "anonymous".

[Yura]

Thank you for advice.
I checked both exponents as advised and this is what is says:

47446111 No factors below 2^68 P-1 B1=565000, B2=17373750 Assigned LL testing to "Yura" on 2010-02-09 History B1=565000, B2=17373750 by "MCAS" on 2009-11-16

47445667 No factors below 2^68 P-1 B1=565000, B2=17373750 Assigned LL testing to "Yura" on 2010-02-09 History B1=565000, B2=17373750 by "MCAS" on 2009-11-17


The point is that those two exponents have been completed AND sent automatically (not manually) and here is corresponding log items from results.txt:

[Fri May 07 02:25:13 2010]
UID: YURA, M47445667 is not prime. Res64: 7652310E94CD____. Wd1: 8D841E2D,934591,00000000, AID: XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX
[Fri May 07 15:02:44 2010]
UID: YURA, M47446111 is not prime. Res64: 68550F47623A____. Wd1: 41115914,10738995,00000000, AID: XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX

As per prime95, results have been sent and now I am processing two new prices and exponent status reports them correctly assigned to me.

So what happened with previous two, finished exponents?

[axn]

can someone mask the residues, please?

[lycorn]

That is not to be expected, indeed.
It might be that the exponents were completed (and hence Prime95 correctly wrote the appropriate lines to results.txt), but due to comms problems they didn´t make their way to the server (and hence they are still assigned to you). Try the following:
1) Log in with your server username (this is important)
2) Go to the Manual Testing -> Results page
3) Copy to the clipboard the 4 lines from results.txt that you have posted in your last post.
4) Paste them in the paste area of the server page you have just accessed, and click the Submit button.

Cross your fingers (kidding...) and watch the messages that appear on the screen. The exponents are now supposed to be considered as tested and credited to you. Check that´s the case and keep us posted. If that fails for any reason, we´ll try another workaround...

[cheesehead]

Originally Posted by Yura

corresponding log items from results.txt:

[Fri May 07 02:25:13 2010]

UID: YURA, M47445667 is not prime. Res64:
<snip> Wd1: 8D841E2D,934591,00000000, AID: <snip>

[Fri May 07 15:02:44 2010]

UID: YURA, M47446111 is not prime. Res64: <snip> Wd1: 41115914,10738995,00000000, AID: <snip>

Did you look in your prime.log file for the reporting transaction? (Have you noticed your prime.log file?)

Search on "47445667" in prime.log. There should be some lines similar to:

Sending result to server: UID: YURA, M47445667 is not prime, Res64: xxxxxxxxxxxxxxxx. Wd1: 8D841E2D,934591,00000000, AID:  xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx

PrimeNet success code with additional info:
CPU credit is nn.nnnn GHz-days.

... if your result was successfully reported and credited, that is.

If there are no such lines, show us any lines in prime.log that are associated with your two exponents (but please mask the residues and AIDs - so someone can't copy them from your posting and use them to submit a fraudulent report).

[Yura]

Thank you lycorn for advice. I did as you suggested and it looks like it worked!!! After pasting my results I got this:

No factor lines found: 0
Factors found: 0
P-1 lines found: 0
LL lines found: 2
Processing result: M47445667 is not prime. Res64: . Wd1: , AID:
CPU credit is 82.3709 GHz-days.
Processing result: M47446111 is not prime. Res64: . Wd1: , AID:
CPU credit is 82.3717 GHz-days.
Mlucas lines found: 0
Glucas (G29) lines found: 0
Glucas lines found: 0
MacLucasFFTW lines found: 0
ECM lines found: 0

To cheesehead; I looked through prime.log but there is not one line similar to your sample i.e. no messages regarding sending results to the server. I believe manual pasting results did the trick and eventually this table will get updated and ‘manual testing’ will be removed:
Manual testing 0 47946991 LL 0.00% 2010-02-09 18:07 106 2010-02-09 18:07 2010-07-29 18:07
Manual testing 1 47947147 LL 0.00% 2010-02-09 18:07 106 2010-02-09 18:07 2010-07-29 18:07
0 50475973 LL S1, 0.00% 2010-05-10 20:51 16 2010-05-10 20:51 2010-05-17 20:51 2010-07-13 14:51 48
1 50476021 LL 0.00% 2010-05-10 20:51 16 2010-05-10 20:51 2010-05-17 20:51 2010-07-13 14:51 48



As for the masking of ‘AIDs’, I thought mathematicians and people interested in mathematics are honest bunch but I might be wrong? What will be the point to claim results which do not belong to you? Probably money and fame is involved…

Thank you guys for your unselfish help!

[lycorn]

Good, I´m glad it´s sorted out!
For some reason the program didn´t communicate with the server, that´s why you didn´t find any trace in the prime.log file.
If you look into your Results page, you´ll see the exponents credited to you, and they will be no more on your Assignments page.
As for the need to mask keys, residues, and similar stuff before the exponent is successfully double checked, well, there are malicious (and/or stupid) people everywhere, so we need to take a pragmatic approach and take some precautions. I don´t believe a mathematician would do that, and it´s not about fame or money either, but more like some stupid joke.
Anyway, congrats on your first results, and welcome to the search. Have fun!

[Mini-Geek]

Originally Posted by lycorn

As for the need to mask keys, residues, and similar stuff before the exponent is successfully double checked, well, there are malicious (and/or stupid) people everywhere, so we need to take a pragmatic approach and take some precautions. I don´t believe a mathematician would do that, and it´s not about fame or money either, but more like some stupid joke.

Yeah, where you're just showing that a Mersenne number is composite there's no money and very little fame involved (being one of GIMPS's highest scoring members may provide a little fame among other math geeks...until they find that it's all by faking results ).
But by masking residues it means that people, whether malicious or stupid, are unable to resubmit someone else's first-time LL result as a matching DC result to claim credit.
Besides preventing credit where it's not due, this means that GIMPS's mathematical results (most importantly the determination of composite or prime, and secondarily of factors and LL residues) have high integrity, so you don't have to second-guess whether GIMPS missed a prime somewhere due to a fabricated/erroneous LL result and frabricated DC result.
The masking of assignment keys is more about preventing poaching and avoiding duplication of work, for better overall efficiency. It lets the users have proper control over what they've been assigned.