Yes, the 255 comes from msieve. This part of the script is very simple so its not likely to delete anything when it is stopped.

Sorry I lost you a weeks work :-(

[quote=Brian Gladman;210188]Sorry I lost you a weeks work :-([/quote]
Actually, I figured I lost it because I wasn't able to take the time to get with you as to the why. And, it was a bonus already, since I took the machine with me and was able to let it run 24/7. I might have lost several weeks had I not taken it along. I do need to see if I saved any info from that crash. I must have - I don't throw anything away.:smile:

And, it's all for fun anyway, right? Why else would we do it?

I thank you again for all your work.

Well, this is what happened this time. I believe this might be the same thing as last time:
[code]
Total yield: 1080071
0/0 mpqs failures, 102757/29996 vain mpqs
milliseconds total: Sieve 8807326 Sched 0 medsched 1016971
TD 4201649 (Init 232386, MPQS 517840) Sieve-Change 1305112
TD side 0: init/small/medium/large/search: 45439 488287 64223 155434 1250169
sieve: init/small/medium/large/search: 133982 2128119 154521 1825986 475410
TD side 1: init/small/medium/large/search: 25129 353424 58108 157481 1369759
sieve: init/small/medium/large/search: 92838 1610179 151621 1935647 299023
appending spairs.out.T0 to spairs.out
appending spairs.out to test.dat
Found 7556726 relations, 110.9% of the estimated minimum (6816250).
Traceback (most recent call last):
File "c:\MathWork\ggnfs\factmsieve.py", line 2004, in <module>
run_siever(client_id, num_clients, NUM_THREADS, fact_p, lats_p)
File "c:\MathWork\ggnfs\factmsieve.py", line 1718, in run_siever
if run_msieve('-t {0:d} -nc1'.format(NUM_CPUS)):
File "c:\MathWork\ggnfs\factmsieve.py", line 325, in run_msieve
rel = os.path.relpath(cwd)
File "c:\python31\lib\ntpath.py", line 591, in relpath
raise ValueError(error)
ValueError: path is on mount 'C:', start on mount 'c:'
siever terminated
WARNING: gnfs failed to find a factor. This really shouldn't happen.
I'll just run ecm till the end of time or a factor turns up...
Let's hope you don't run out of disk space before either of those.
[/code]
Is this a Python problem? Could there be an issue of case sensitivity, even though this is Windows?

Is this recoverable? I couldn't find a way last time. Thanks for any help.

OK, I have changed all the references to upper case and restarted the script in hopes it doesn't overwrite something. If it does, not really a huge loss...

I should know something later today...

The restart appears to have functioned well. Other than the extra sieving range that ultimately took the relations to 142%+, all else seems to be proceeding along.

Another tidbit learned - Python is case sensitive, even in Windows...

I tried the version 0.62, but this one seems to have trouble with spaces in the GGNFS/MSIEVE_PATH even though I'm using the 8.3 filenames. I haven't had the time to look into this, but 0.51 worked fine. What did change in between?

[quote=smh;210514]I tried the version 0.62, but this one seems to have trouble with spaces in the GGNFS/MSIEVE_PATH even though I'm using the 8.3 filenames. I haven't had the time to look into this, but 0.51 worked fine. What did change in between?[/quote]

There are two changes in path handling between these two versions:
(1) On Linux './' is added to the front of the msieve path.
(2) on both Windows and Linux, msieve is run in its own working direcory with other paths being adjusted accordingly. The normal working directory is restored when msieve completes.

I have looked at these and I don't see any obvious differences in the handling of spaces. Are files not being found or are there specific Python error reports?

Brian

Interesting, the C: vs. c: trouble I had with 0.62 never cropped up with 0.51. Perhaps related to the spaces issue?

In thinking about it, the only change between 0.51 and 0.62 on my systems, was the script file itself. So, there didn't seem to be any case sensitivity with 0.51. Therefore, I can't see it being something with Python directly, since that didn't change...

BTW, 0.62 now appears to be running fine here on both linux and WinXP machines.

Thank you for all the work.

I did try to reproduce the problem on my laptop (which worked fine with 0.62).

Aliqueit is running in "D:\factor\ali quot\aliqueit" (I simply added a space in aliquot) and changed aliqueit.ini accordingly [CODE]//Name (and path, if needed) of your ggnfs executable.
ggnfs_cmd = D:\Factor\GGNFS\Python\App\python.exe D:\Factor\ALIQUO~1\aliqueit\factmsieve.py[/CODE]

[CODE]appending spairs.out.T0 to spairs.out
appending spairs.out.T1 to spairs.out
appending spairs.add to spairs.out
appending spairs.out to test.dat
Found 138171 relations, 276.3% of the estimated minimum (50000).
-> msieve -s ..\..\Ali quot\aliqueit\ggnfs_74025328271242008403164217851901317606436418593662363728613970318072695126489660406944677\test.dat -l ..\..\Ali quot\aliqueit\ggnfs_74025328271242008403164217851901317606436418593662363728613970318072695126489660406944677\test.log -i ..\..\Ali quot\aliqueit\ggnfs_740253282712420084031642178519013176064364185936623
63728613970318072695126489660406944677\test.ini -nf ..\..\Ali quot\aliqueit\ggnfs_74025328271242008403164217851901317606436418593662363728613970318072695126489660406944677\test.fb -t 2 -nc1
cannot open input file '..\..\Ali'
compressing spairs.out to spairs.save.gz
-> making sieve job for q = 350000 in 350000 .. 355000 as file test.job.T0
-> making sieve job for q = 355000 in 355000 .. 360000 as file test.job.T1
-> Lattice sieving algebraic q from 350000 to 360000.
-> gnfs-lasieve4I11e -k -o spairs.out.T0 -v -n0 -a test.job.T0
-> gnfs-lasieve4I11e -k -o spairs.out.T1 -v -n1 -a test.job.T1[/CODE]It is now looking for '..\..\Ali'

I've tried the full path between quotes and also the short filename.

smh, could you try changing the following two lines in version 0.62 of the script?

Change line 328 from:
[CODE]
args = ('-s {0:s} -l {1:s} -i {2:s} -nf {3:s} '
[/CODE]
To:
[CODE]
args = ('-s \"{0:s}\" -l \"{1:s}\" -i \"{2:s}\" -nf \"{3:s}\" '
[/CODE]

And, change line 2030 from:
[CODE]
args = ('-s {0:s} -l {1:s} -i {2:s} -v -nf {3:s} -t {4:d} -nc3'
[/CODE]
To:
[CODE]
args = ('-s \"{0:s}\" -l \"{1:s}\" -i \"{2:s}\" -v -nf \"{3:s}\" -t {4:d} -nc3'
[/CODE]

[QUOTE=WraithX;210551]smh, could you try changing the following two lines in version 0.62 of the script?[/QUOTE]

Actually, I should have tried that before posting it. It does not work on my machine. My apologies for leaping before looking. I thought it would work, but I get the following error with that change:
[CODE]
-> msieve -s "..\c123\c077.dat" -l "..\c123\c077.log" -i "..\c123\c077.ini" -nf
"..\c123\c077.fb" -t 3 -nc1
cannot open input file '"..\c123\c077.ini"'
[/CODE]

This won't be as easy to fix as I thought. Hopefully the architect will know better how to proceed. :smile: