Programming and Scripting :: C question



After running one process with the fix for two days, I've come to conclusion it works great.
So, I applied it to all, and restarted.

Load averages are only 12 now, so it has a good speedup effect.
Glad it's working out :=)

One inefficiency is it is generating some perms more than once - because  we want the module to allow repetition of chars the same perm can come up again.  The module's API doesn't seem to have a way to prevent that.  However, that's probably outweighed by the speed gains of using the module Cf. Perl code from scratch.

I'm sure there would be more efficient, optimized, purpose-designed cracking progs written in Assembly or something that avoid generating the same perm more than once etc.

In any case, there's still the main holdup - that of trying every combination on openssl, which far outweighs other bottlenecks.

There may be a faster perm Perl module around than this or some Perl programmer has very probably written a better approach - I might have a look.

Regardless, this is still infinitely better than trying to do it by brute force in bash.  Perl is much better suited to this type of task.

EDIT1:Thought I'd bookmark this link which discusses migrating a similar type of simple prog from badly-written Perl to much faster C:

http://blade.nagaokaut.ac.jp/cgi-bin/scat.rb/ruby/ruby-talk/203958

EDIT2:  This is interesting reading about the (very different) word list approach used in  "Crack" by its author:
http://www.crypticide.com/dropsafe/article/733
Status report. They have been running for 46 days nonstop now. They have punched through about 5 million combinations. This is nice, and still not cracked.

About a week ago I ordered a new comp for me; I feel the need for more power :p
It's a huge bang-for-the-euro combination. Specs:
Core 2 Duo E2160 @3Ghz
2GB Kingston ram 667Mhz (2x 1gb sticks in dual channel)
160gb HD
And, as I'm not a gamer, didn't waste my time with a graphics card. I will put in one of my pci ones, most likely the best, Ati Rage II+dvd.
Total price 267 euros.

It's a really powerful one, hopefully more like 8x my current best one (this P3). As with all my other comps, I will build the OS for it, to gain as much speed as possible. Also this is my first 64-bit capable machine, so I will make the OS fully 64-bit, to get 10-20% more power for some apps. Plus optimize boot speed, and load parts to ram, I like my operating systems blazing :D

So, when it arrives, and I finish the OS, this cracking will move onto it. Despite having to start over, I'm quite confident that it, with 2 brutal cores, will take my pass down. It's just a matter, and question, of time.
Wow, your certainly serious about this project!  New machine sounds *nice*.

I suspect that to do permutation with replacement without producing redundant perms we're supposed to use the Mathematics::Combinatronics module or something.

Best that I either get some help on Perlmonks (probably a good idea), or we work out how to write this in C.  I do have access to an A#1 world guru C/C++ programmer at work who might help me if I get stuck.  Let's see.

How long before your new box comes onstream?
It will hopefully be delivered next week *sigh*, as they have all parts but the mobo is backordered.

This comp is for other tasks also, I couldn't have afforded a new one just for this :D

Oh, I use 10 modified versions of your previous perl core, and 1 of the new one with the Permutate module (because the prev ones' starting point could be modified). I agree writing this in C might bring great speed gains also.

Also, the 5 million was just a guess, as my log file only includes *skipped* ones :D So hopefully they have gone through more.
Next Page...
original here.