hashcat Forum

(06-04-2012, 12:12 PM)Pixel Wrote: [ -> ]Well done Hash-IT, I like this idea. Would it still work after filtering with sed?

Look like I'll be getting that new GPU after all....

Thanks !! It just came to me in a flash of inspiration ! I don't get those very often !

However, M@LIK seems to think it is too simplistic so just hold on a while to see if he or anyone can point out where I have got it wrong !! I don't doubt I have messed up somewhere.

Yes, please get yourself a GPU, you will love it. There are some very cheap secondhand HD5970's about just now and to be honest there isn't much faster for the money.

I wish I had more than one !!! I can get 99K wpa with atom's improvements to v0.8.

could some one pls explain to me when you use rule does it run CPU or GPU to modify the dict? also when it breaks up with a rule and go to the next? When the rule can not find a match or ...?

(06-05-2012, 02:08 AM)ntk Wrote: [ -> ]could some one pls explain to me when you use rule does it run CPU or GPU to modify the dict? also when it breaks up with a rule and go to the next? When the rule can not find a match or ...?

Using rules in ocl-hashcat-plus uses GPU. atom said it actually makes full use of your GPU when reading from a wordlist to add some rules as it is much more efficient.

When there is no match for a rule it is ignored.

thanks

You're welcome.

Have you had any results testing the pattern + rule theory out ? I am worried about it since M@LIK mentioned it my not be that simple. However I haven't thought of a problem with it yet, have you ?

Sorry I wasn't clear enough when I posted the rule list to make this work. I just generated them and in my haste pasted them in a column instead f in a line. Here are the rules as they should be.

The ###### is just to separate them for clarity, don't add that to the rule !

This idea seems to work well Hash-IT but we need to find faster way to filter it and figure how hard to filter it.

(06-07-2012, 12:27 PM)Pixel Wrote: [ -> ]This idea seems to work well Hash-IT but we need to find faster way to filter it and figure how hard to filter it.

I'm glad you are still interested in this idea

Yes I think the "mask" idea works well in theory, I can see other uses for this also, perhaps hex !!

Unfortunately the rule feature "s" doesn't work in the way I originally though it did. I wrongfully assumed that you could multi re-map all characters but you cannot, only one at a time.

This will explain why M@LIK didn't like the idea.

I have made a feature request for multi re-mapping of characters to be available in ocl-hashcat-plus. This is the only way forward for this project I think. If we can persuade atom to implement this feature then we can have fun making new optimised masks for all sorts of things.

Got another idea for only needing to generate 1 set of characters, instead of swapping/replacing can't we just generate the full A-Z Upper length 7 and just Insert the character that's missing with a rule like this?


Ive generated 26^7 and its 67.3GB so its abit smaller too.

---

EDIT
Ive just generated 26^6 and its only 2.3GB could filter this with sed and Insert the first and last characters that are missing.

Guessing this will effect the overall output as sed won't know what the Inserted characters are.

here's the rule for 6 characters or 26^6

Hi there people.
I am new here.
A brief introduction , i read Pixels post in Tapes blog searching for filtering wordlists and driven from pure scientific interest i came here spending a few hours reading this whole lovely thread. I am a math teacher myself but i have basic understanding of programming and putting an end to introduction i d like to state my humble opinion.
I feel many people could help here but since we don't have the level of expertise with sed or other filterers it would be nice "if u like it of course" to drive us through basic level knowledge. I don't represent laziness here, on the contrary i i believe in hard work but its impossible to travel through the vastness of knowledge without expert guidance. And u seem willing driven and expert to guide many like me around the filter processing.
About my motive: in my country there is a vast majority of routers that use 12 digit passwords 6 of whom are broadcasted (known) with the bssid and 6 which are obviously hidden. These are lower alpha+ numeric which means 36**6=around 2.2billion passwords around 22gigabytes. Now as you probably know i would like to filter something like that, like you do here, which is really close to my case and significantly reduce data and time.
I will be spending next few days with sed self tutoring but any help would be welcome. If i understand all the commands 100% i might be able to help u with filtering. And when it comes to calculation power i volunteer to take up one letter for you when u reach your final algorithm and start sharing duties.
I respect what you do, the free open source and common spirit.
cheers