![]() |
![]() |
#1 |
Oct 2008
Annapolis, Md
7 Posts |
![]()
Hi All,
I have a Dual core Dell running V5 of the Prime95. I have checked my web results and they show that I have 2 exponets assisgned (1 per core) and show 1 at 99.7% complete and another at 0% complete. When I look in my results.txt file I find that they both have been completed, not prime ![]() Any direction would be helpful. ed "I'm in the Prime of my Life" |
![]() |
![]() |
![]() |
#2 |
P90 years forever!
Aug 2002
Yeehaw, FL
2·23·173 Posts |
![]()
Email the prime.log file to me so I can check it for anomalies.
Then copy your results.txt output to the manual results web page at http://www.mersenne.org/manual_result/ |
![]() |
![]() |
![]() |
#3 | |
1976 Toyota Corona years forever!
"Wayne"
Nov 2006
Saskatchewan, Canada
3·37·47 Posts |
![]() Quote:
Did you try Manual Communication? |
|
![]() |
![]() |
![]() |
#4 |
"Richard B. Woods"
Aug 2002
Wisconsin USA
22×3×641 Posts |
![]()
I have a similar problem with v25.8.
Short description: I've never seen a prime.spl file in the Prime95 directory for v25.8 since I began using it a couple of weeks ago. There's a prime.log and prime.txt, but no other prime.* or *.spl file. More details, in case #1) isn't the key symptom: I upgraded to prime95 v25.8 a couple of weeks ago on an XP system. Since then, I've kept UsePrimenet=0 most of the time, but turned it on and used Manual Communication to get some assignments. Other times, I've gotten assignments via the manual_assignment page. I used only the manual_result page to report results ... until today. I dialed in, then turned on UsePrimenet and Manual Communication. (DaysOfWork was 5 and DaysBetweenCheckins was 7.) v25.8 got several new assignments and reported expected completion dates, but apparently didn't report any of the several completed TF and P-1 assignments I had not yet reported. I then changed DaysBetweenCheckins to 3 and retried Manual Communication. Again, no communication of completed results. After it finished communicating, I checked the exponent statuses. None of the assignments that I hadn't already reported manually in the past showed that my work had been done. Exponents for which I completed TF from 2^64 to 2^65 (as shown in results.txt) still were displayed as having been TFed only to 2^64. (I can report results on the manual_result page, of course, but that doesn't solve the problem.) Last fiddled with by cheesehead on 2009-01-24 at 07:48 |
![]() |
![]() |
![]() |
#5 |
P90 years forever!
Aug 2002
Yeehaw, FL
2·23·173 Posts |
![]()
When an assignment completes and UsePrimenet is set to 0, then the client doesn't think it needs to send the result to the server. It will write the result to the results.txt file and not create a prime.spl file. In this case, the only way to record the result is to use the manual results web page.
|
![]() |
![]() |
![]() |
#6 | ||
"Richard B. Woods"
Aug 2002
Wisconsin USA
170148 Posts |
![]() Quote:
Quote:
"Error code: 40, error text: TF result for Mxxxxxxxx was not needed ... (repeated for each TF) ... Error code: 40, error text: P-1 result for Mxxxxxxxx was not needed" ! Sure enough, the exponent status report now shows those results. So, what happened? While I was messing around with UsePrimenet, did the client "change its mind"? Does it not need the .spl file? Can it report straight from results.txt? ![]() Or did it just happen that there _was_ an .spl file, but not at the particular times when I looked -- somehow I just happened to miss looking for it when it did exist? ![]() Edit: Just looked again at prime.log. Found _two_ "Sending result to server" lines, but each time (once on Jan 18, once earlier this morning) there's only one result mentioned, not the other 18-19 TFs also done (and _now_ showing as done on the exponent results report. Again: the prime.log shows only 1 result being reported this morning, but the exponent results report shows all 20 results and has a 2009-1-24 date on all of those. Here's a guess: The two results shown in prime.log as being sent were both TFs that just happened to complete during the time that I (a) had UsePrimenet=1 and (b) was logged-on. None of the other TF or P-1 completions occurred when those two conditions were true. More breaking news Right now, there is a prime.spl file! Timestamped same minute as prime.txt (which has UsePrimenet=0), but two-and-a-half hours after the prime.log timestamp. Content (16 bytes) (I can't figure out how to make TextPad display it in hex): ¬’Ós(another char -- see below)(null)(null)(null)@(null)(null)(null)d(null)(null)(null) There's a character between the "s" and the first null that doesn't show up after I save the edit, but when I do a copy-and-paste here it looks like a four-spotted "domino" but instead of spots has "0 0" on one row and "0 1" on the row beneath it. Last fiddled with by cheesehead on 2009-01-24 at 17:46 |
||
![]() |
![]() |
![]() |
#7 |
P90 years forever!
Aug 2002
Yeehaw, FL
2×23×173 Posts |
![]()
OK, there is a way to get a prime.spl file when UsePrimenet=0, but it is a minimum size that just says either "ping server", "send new end dates", or "update computer info".
However, the key routine that spools results is SpoolMessage and the very first line of that routine is "if (!USE_PRIMENET) return;". |
![]() |
![]() |
![]() |
#8 |
Aug 2002
Termonfeckin, IE
24·173 Posts |
![]()
Why would you want to set UsePrimenet on and off? Why not just set UseManualComm=1. That way you have control over when the machine contacts Primenet?
The only reason I would see to do what you are doing is if you want certain exponents to be worked on "outside" of Primenet and certain others "inside" Primenet. In that case, you need to pay attention to when the flag is set and unset as George indicated. Last fiddled with by garo on 2009-01-26 at 12:24 |
![]() |
![]() |
![]() |
#9 |
"Richard B. Woods"
Aug 2002
Wisconsin USA
22·3·641 Posts |
![]()
What would be the consequences of removing that test and adding some way for manual reporting to reset the prime.spl file?
Last fiddled with by cheesehead on 2009-01-26 at 17:02 |
![]() |
![]() |
![]() |
Thread Tools | |
![]() |
||||
Thread | Thread Starter | Forum | Replies | Last Post |
Reporting partial ECM results? | apsen | PrimeNet | 3 | 2011-07-03 14:54 |
results not reporting | Unregistered | Information & Answers | 2 | 2010-05-23 00:56 |
Manual reporting of non-Prime95 results to V5 | Graff | PrimeNet | 5 | 2008-08-18 11:58 |
Reporting Results | dave_0273 | Marin's Mersenne-aries | 2 | 2005-10-18 12:54 |
Problem with Reporting Results to Primenet | RMAC9.5 | Software | 3 | 2003-11-02 23:06 |