04-05-2015, 10:13 PM
I deleted my original post after finding the wiki page that details the 31 character limit, but decided to repost it... it's between the dashed lines. But I have other questions/issues...
-------------------
I'm doing testing now that I got my GPU's working, and it won't crack a password. I know it's in there as I SHA1'ed the password myself and pasted it in the command line.
When I SHORTEN the password, it works.
When i crack this: 903b8b9cbd54b01dce32d0038821a7725c97806a
Which is: 8|7|9|4|2|10|6|11|12|1|0|5|13|3|14
It does not find it.
When I crack: 4845d07488575413261631f67d57e5678c5ccc8b
Which is: 8|7|9|4|2|10|6|11|12|1|0|5|13|3|1
It finds it. It finds a length 33, fails on 34. Huh?
Now I know old versions of oclhashcat had a limit of 15, but nowhere am I seeing a limit of less than 55 (found that googling somewhere).
Anyone have any idea?
---------------
So, like I said, I found the page that says it can crack a 55 character password, but, it has to be in 2 parts, and each part cant be over 31 characters. That SUCKS. I actually bought $800 in hardware today for this project I'm working on, only to find out I can't do it because of that odd restriction? To top it off - closed source - I can't just change the code (32 characters sure sounds like a bit-related limit to me).
I cant even use the combiner - I'm working with a permutation of numbers. I may have to try it with a special set of masks. Which will VAAAASTLY increase the cracking time. 13 numbers, 0-12, no repeats:
Using my permutated list:
13x12x11x10x9x8x7x6x5x4x3x2x1=6,227,020,800 possible options
Using the masks, assuming I can set a 0-12 range on a single 'character':
13x13x13x13x13x13x13x13x13x13x13x13x13=302,875,106,592,253 hashes
Forty eight THOUSAND times the work.
And this list of 13 numbers is a preliminary test, I was hoping to work with numbers up to 18 (which is absurdly high, but still reasonable if it's a permutation).
Is there ANY way to do this, or am I screwed by a programming limitation?
Can anyone think of a fancy way to make it work? I briefly thought of tables where it was split in half, but that multiples the work many times over as well.
Ugh
-------------------
I'm doing testing now that I got my GPU's working, and it won't crack a password. I know it's in there as I SHA1'ed the password myself and pasted it in the command line.
When I SHORTEN the password, it works.
When i crack this: 903b8b9cbd54b01dce32d0038821a7725c97806a
Which is: 8|7|9|4|2|10|6|11|12|1|0|5|13|3|14
It does not find it.
When I crack: 4845d07488575413261631f67d57e5678c5ccc8b
Which is: 8|7|9|4|2|10|6|11|12|1|0|5|13|3|1
It finds it. It finds a length 33, fails on 34. Huh?
Now I know old versions of oclhashcat had a limit of 15, but nowhere am I seeing a limit of less than 55 (found that googling somewhere).
Anyone have any idea?
---------------
So, like I said, I found the page that says it can crack a 55 character password, but, it has to be in 2 parts, and each part cant be over 31 characters. That SUCKS. I actually bought $800 in hardware today for this project I'm working on, only to find out I can't do it because of that odd restriction? To top it off - closed source - I can't just change the code (32 characters sure sounds like a bit-related limit to me).
I cant even use the combiner - I'm working with a permutation of numbers. I may have to try it with a special set of masks. Which will VAAAASTLY increase the cracking time. 13 numbers, 0-12, no repeats:
Using my permutated list:
13x12x11x10x9x8x7x6x5x4x3x2x1=6,227,020,800 possible options
Using the masks, assuming I can set a 0-12 range on a single 'character':
13x13x13x13x13x13x13x13x13x13x13x13x13=302,875,106,592,253 hashes
Forty eight THOUSAND times the work.
And this list of 13 numbers is a preliminary test, I was hoping to work with numbers up to 18 (which is absurdly high, but still reasonable if it's a permutation).
Is there ANY way to do this, or am I screwed by a programming limitation?
Can anyone think of a fancy way to make it work? I briefly thought of tables where it was split in half, but that multiples the work many times over as well.
Ugh