Safe to delete prime.spl file?

[Graff]

Is it safe to delete the prime.spl file in order to prevent Prime95 trying to
unreserve an exponent that was not assigned to me?

Here's the situation. After installing a MS Vista patch about a week ago
my choice of screen resolutions shrank (leaving me with a max of
1280x1024). I did a system restore, which restored all the previously-
available screens, but (contrary to statements like "System restore will
not affect userfiles"...) buggered up Prime95, causing it to start doing
already-completed exponents upon post-system-restore boot.

After having to reboot this morning at the request of the McAfee package,
I was again confronted with missing screen modes! After getting back
from work, I did another system restore. Again, Prime95 files were
replaced with older versions. After a lot of fiddling, I have the machine
working on the exponents it was working on before. I have lost a day
or so on the first-time LL test and, just to cause more grief, Prime95
is repeating the 2560K FFT test as the current exponent exponents is
the first one to require this size.

So my question about deleting the prime.spl file is prompted by the
fact that five new factoring assignments were grabbed after the latest
system restore and the five factoring assignments I had previously were
missing from the worktodo.ini file. I replaced the missing assignments
at the top of the worktodo.ini file and then used the "unreserve exponent"
to return the five newly-assigned exponents. Unfortunately, I mistyped
the second exponent. So every 15 minutes it tries to unreserve that
exponents, barfs and then fails to unreserve the next three.

Can I simply delete the prime.spl file?
Can I then simply unreserve the four exponents again? Or will I have
messed up the Prime95 system?

Graff

[RMAC9.5]

Yes, in your case it is safe to delete the prime.spl file. Prime95 uses this file to temporarily store the text messages that it is going to send to the server at its next communications opportunity. Prime95 deletes this file when it successfully communicates with the server and creates a new one when it has more stuff to send. Based on your description of the events that have transpired, the prime.spl file contains stuff about one or more of the factoring exponents that you are trying to unreserve. Why not just keep the newly assigned exponents?

[ric]

Graff,
you might also want to unreserve manually one (or more) exponent at a time at:

http://mersenne.org/ips/manualtests.html

[Graff]

Originally Posted by RMAC9.5

Yes, in your case it is safe to delete the prime.spl file. Prime95 uses this file to temporarily store the text messages that it is going to send to the server at its next communications opportunity. Prime95 deletes this file when it successfully communicates with the server and creates a new one when it has more stuff to send. Based on your description of the events that have transpired, the prime.spl file contains stuff about one or more of the factoring exponents that you are trying to unreserve. Why not just keep the newly assigned exponents?

To answer your last question, because the newly-assigned exponents
already seem to have been factored to 2^68. I recall something about
new ranges for factoring limits, so these exponents may have needed
trial factoring to 2^69, but given the hassles I was already fixing, I didn't
want to investigate that. It was just simpler to unreserve them.

Graff

[Graff]

Originally Posted by ric

Graff,
you might also want to unreserve manually one (or more) exponent at a time at:

http://mersenne.org/ips/manualtests.html

I ended up doing that.

I also reread the README and UNDOC files to see whether I was missing
something obvious. I didn't see anything that might help.

I ended up downloading the source code and examined the code to see
what it was doing with the PRIME.SPL file. I found some code in the
"communicate results to primenet" section that would ignore the
ERROR 3 messages after 72 attempts. So I set the network retry
time to 1 minute, went upstairs to watch a DVR'ed episode of
Mythbusters, then came downstairs to find that the 72 attempts
had been made, that the next attempt had successfully sent the
contents of the PRIME.SPL file (including one no-factors-found
and one factor-found result) and everything seems back to normal.

I think the next time I have to do a system restore, I'll ensure that I
make a backup of the Prime95 directory beforehand so that I don't
lose any work.

Graff