PDA

View Full Version : Retrieving work in linux?



strates
01-27-2006, 08:27 PM
Hello everybody,

I would like to start a discussion on an effective way to retrieve proth tests lost while using the linux client. Similiar to the how-to for windows.

Thus far, I have opened the cache file in a hex editor, and attempted to edit the applicable bytes with the correct test information using little endian encoding. However when I reloaded the program, it told me the checksum was bad on the cache and downloaded a new test.

I guess the discussion should involve which bytes in the cache need to be changed along with the standard k and n bytes, so that the checksums remain valid, and the client is tricked into loading the correct test.

as an example I will show you the changes I made in my first attempt. The test I wanted to reload had the same k value, so I only changed the first 3 duplets. My client was working on n=6674671, which is equal to 0065D8EF in hexadecimal, it is stored in the cache file though, in little endian encoding, so it was stored as EF D8 65 00. The test I wanted to save was n=5029711, which is equal to 004CBF4F, and so I edited my cache file so that the first 4 duplets were now 4F BF 4C 00.

Upon reloading the sb program, I recieved the checksum error, and started a new proth test.

Anybody able to see what I did wrong right off the bat?

Thanks in advance,
Strates

Ken_g6[TA]
01-30-2006, 12:51 PM
I don't know how the checksum works, but I've written a program (http://home.att.net/~k.brazier/programs/java/sbserver/) (ServeKN.java) that would let you fix your problem.

engracio
11-11-2007, 01:14 PM
Hello everybody,

I would like to start a discussion on an effective way to retrieve proth tests lost while using the linux client. Similiar to the how-to for windows.

Thus far, I have opened the cache file in a hex editor, and attempted to edit the applicable bytes with the correct test information using little endian encoding. However when I reloaded the program, it told me the checksum was bad on the cache and downloaded a new test.

I guess the discussion should involve which bytes in the cache need to be changed along with the standard k and n bytes, so that the checksums remain valid, and the client is tricked into loading the correct test.

as an example I will show you the changes I made in my first attempt. The test I wanted to reload had the same k value, so I only changed the first 3 duplets. My client was working on n=6674671, which is equal to 0065D8EF in hexadecimal, it is stored in the cache file though, in little endian encoding, so it was stored as EF D8 65 00. The test I wanted to save was n=5029711, which is equal to 004CBF4F, and so I edited my cache file so that the first 4 duplets were now 4F BF 4C 00.

Upon reloading the sb program, I recieved the checksum error, and started a new proth test.

Anybody able to see what I did wrong right off the bat?

Thanks in advance,
Strates


Is there any update howto on this subject. I lost one k/n pair and the box grabbed another wu. It is running Fedora 6. I know we have an excellent howto for windows which I've used dozen of times moving wu around. Seems like there is no Linux version. Thanks.

Joh14vers6
11-11-2007, 03:24 PM
']I don't know how the checksum works, but I've written a program (http://home.att.net/~k.brazier/programs/java/sbserver/) (ServeKN.java) that would let you fix your problem.
How can I run this program?

engracio
11-12-2007, 11:50 PM
I guess if you use Linux you're SOL if you get a segfault and 95% done on a wu. Poop start a new one.:rotfl:

I understand Linux don't do it as much as Windows:thumbs: coolest. Can I deal? Yes I can.:Pokes: