A problem with worktodo.txt

Hello:

I'm a new user and I have a problem with worktodo.txt file. The program mprime told me that the format was wrong, and I want to know where is the mistake in order to process properly.

I'm doing all the computations offline. When I asked for tasks, the server brings me the next message:

====================================================================

Test=8B1614264D99F0C8C36334941AB46621,332227993,75,0
Test=D1E406341A1A749702761E2A46664FAD,332219897,73,0

Distribute the lines above to your computer's worktodo.txt files. A typical version 25 prime95 worktodo.txt file on a dual core computer looks like this:

[Worker #1]
Test=A5996856A4F73D28634EFA6CB510066F,exponent,68,1
[Worker #2]
DoubleCheck=B94D34BF558295F697581568AD2E14AF,exponent,66,1

Older prime95's must run one prime95 instance for each core. The input file is called worktodo.ini. A typical worktodo.ini file looks like this:

Test=exponent,68,1
or
DoubleCheck=exponent,66,1

For Glucas and Mlucas follow the documentation that comes with those programs

====================================================================


I have a dual core, so I want to take that advantage (otherwise is trashing with the 50% capacity of my computer). I have the next worktodo.txt:


====================================================================

[Worker #1]
Test=8B1614264D99F0C8C36334941AB46621,332227993,75,1

[Worker #2]
Test=D1E406341A1A749702761E2A46664FAD,332219897,73,1

====================================================================

The program only process the second test. Why? Please, help me.

P.D.: I have now the second worker working, and I don't want to restart it.

[petrw1]

Check the menu under:
Test
Worker windows...
I have bolded the important parameters that can affect your situation.

Number of workers to run: 2
Priority: 1
Worker Specific Options

The next three parameters allow you to set each worker to do different or the same work if you choose.

Worker Number: 1,2 or All
Type of Work to get: Several Choices
CPU Affinity: Either "Smart Assignment" or "Run on Any CPU" should be fine.

Multithreading (CPUs to use): 1

[S485122]

Originally Posted by fefu

When I asked for tasks, the server brings me the next message:

Test=8B1614264D99F0C8C36334941AB46621,332227993,75,0
Test=D1E406341A1A749702761E2A46664FAD,332219897,73,0

...

[Worker #1]
Test=8B1614264D99F0C8C36334941AB46621,332227993,75,1

[Worker #2]
Test=D1E406341A1A749702761E2A46664FAD,332219897,73,1

Petrw1 answered your questions, I have a remark though.

You changed the last parameter of each line from "0" to "1". This means that you skip an important part of the testing : a P-1 factoring attempt. Unfortunately the example is to specific and partially wrong since the example of the received assignment contains a specific exponent and the "0" parameter, then the example of the worktodo contains the word "exponent" for exponent and the "1" parameter. ("0" P-1 has yet to be done, "1" P-1 has already be done.)

Jacob

[cheesehead]

Originally Posted by fefu

Test=8B1614264D99F0C8C36334941AB46621,332227993,75,0
Test=D1E406341A1A749702761E2A46664FAD,332219897,73,0

< snip >

[Worker #1]
Test=8B1614264D99F0C8C36334941AB46621,332227993,75,1

[Worker #2]
Test=D1E406341A1A749702761E2A46664FAD,332219897,73,1

I agree with everything Jacob said. Just want to add a couple of things:

When you changed the assignment lines (the "0" -> "1"), you abandoned part of what PrimeNet assigned you. Please stop and exit mprime, put the assignment lines back the way they were assigned to you, then restart mprime.

(AFAIK, your partial LL work will not be wasted -- after the P-1, the LL test will resume from where you stopped it, as long as you don't delete the LL save file.

However, I may be wrong about this -- can someone else verify?)

[Mini-Geek]

There is the possibility that he already did the P-1, (seems unlikely due to the problems in getting mprime to understand the worktodo.txt correctly) or is running it on another computer (seems quite possible, and could be convenient if this dual core doesn't have much memory to give it, and another machine does). In any case, I agree that it should be done before the LLs.

Originally Posted by cheesehead

(AFAIK, your partial LL work will not be wasted -- after the P-1, the LL test will resume from where you stopped it, as long as you don't delete the LL save file.

However, I may be wrong about this -- can someone else verify?)

This is accurate, if you have Prime95 set to keep at least two save files. Just in case, you might want to back up your save file first.
When I just tried starting P-1 with only a save file for LL (starts with 'p') for the exponent present, it saw that it wasn't valid for P-1 and deleted it, (leaving the automatic backup) and saved its checkpoint to a save file for P-1 (starts with 'm'). When I started the LL again, it recovered from the backup LL save file. If you don't have a backup, starting a P-1 with only an LL save file will most likely erase your only LL save file.

[cheesehead]

That's in mprime, right? Not just prime95?

[Mini-Geek]

Originally Posted by cheesehead

That's in mprime, right? Not just prime95?

I'm not able to check that, sorry. My tests were with Prime95 25.11. In any case, making a backup of your LL save file just in case would be prudent.
They do seem to be rather similar though. And I know that the save files are compatible, so I'd expect their handling of save files (e.g. in this scenario) to be very similar, if not identical.

Hello:

Thank you to all. You gave me some ideas to retry it and told me about an important error (the 1 at the end of the lines). I must had told some more information, i.e. I'm using mprime under Ubuntu (Linux) running in an AMD dual core proccesor... excuse me. Well, I changed both the lines of worktodo.txt text file and the work to take from merssene (here was the another problem). It is neccesary to order it despite I execute the program offline.

Well, now my two cores are working fine. Thank you again and have a nice day.

Fefu.