unforum

[Scott]

Yay!!! Seems to work fFine in win32. whee

So like, it's pre-built to use the given puzzle-card text? All i have to do is pick a range and let it go? sweet!

I believe, then, we should have people sign up fFor work units on the wiki. Essentially just a sign up sheet to say "I have checked 01 thru ff", so as to avoid too much duplication of effort.

[Scott]

(stated in chat; reposting cos it's good fFun data)

was just working out some numbers on the rccypt tool. 18,446,744,073,709,551,615 possible keys. assuming work units are 16777214 keys long (which should take a half an hour to work through), That breaks into 1,099,511,758,848 work units. Which works out to 62,757,520 years.

suddenly, having people track theeir work units doesnt seem so important. I know people have been saying it would be very very very long and slow and take countably infinite hours to do. But seeing it in action ... man.

Oh!! Good news!! I was off on my calculations!! I didn't account fFor leap Years! it would only take 62,714,565.3 years!!!! **whew** what a relief!

[snowchyld]

hmm

might be an idea to have it outputting a text file with the work units / results in it so we can maybe merge it somewhere ?

[solitair]

Would it not be an idea to have this with purely alphabetic or alphanumeric keys to cut down the search space. That is assuming that we believe the key will also be alphabetic / alphanumeric?

I'd have a go myself, but the wonderful world of modern corporate software has eroded my c / c++ and evolved it into c# (shudder) to a level where I'm not even going to attempt this, simple as it should be. I also don't have a decent dev environment set up on any of my linux boxes.

OK, enough lame excuses on my part, any other thoughts on cutting down the key search space?

[Langley Moor]

To be honest, I'd be comfortable cutting it down to entirely alphabetic to start with. Would it be possible to load in a dictionary file and remove all words other than 8 letter ones? Just a thought, probably pie in the sky

[BrianEnigma]

I ran it against all the 8-letter words in the standard Linux dictionary yesterday and got nothing.

[firebird]

I'm sure someone has tried this already .... but

Could you try the key

thirtenthlabour (16 characters)

(or variations thereof )

We have been told all the info we need is on the card.

[WolverineFan]

It's possible the key could be something less than 8 letters, or the first part of a word longer than 8 letters. For a shorter than 8 letter word you'd probably pad the start or end with 0's to make it 8 "characters" long. It's also quite likely whatever "word" they used wouldn't be part of a dictionary.

I do think that limiting our initial search space to ASCII-printable character values and 00 makes a lot of sense though. If I'm doing my math correctly (and I'm probably not) that cuts our search space down to 1,677,721,600,000,000 keys. If we use Scott's work units of 16,777,214 keys that takes us down to about 5707 years. Give or take.

At least planet Earth will still be around then

If we can spread the load out among 28,000 machines we can finish in 3 months Maybe we can have Seti@Home people run this for a while? I mean seriously, what's more important here?

[firebird]

Oops ... I clearly misunderstood.

Not 16 characters thirteenthlabour

But 8 characters so 13labour

As good as anything to try ...
(IMHO PxC is clearly expecting us to be able to guess the key ... )

[neuromancer]

Great stuff Bri, cheers!!

I'll shelve my own development of a windows version then...I'm willing to help work on a distributed computing setup. Alternatively I can help build a Graphical Interface for Bri's app, for all those GUI whores out there!!!

Following the Seti@Home model would be good, but it does require some organisation to colate all the data accurately. Basically what I'm trying to say is that the more developers we have, the quicker we can build this thing.

Having said that, I'm sure that the key would be something related to PXC, rather than being an ultra-strong alpha-numeric key; surely MC can't be that evil, can they?

[Gibbet]

What output should one get from the prog, assuming the unlikely event of guessing something that gives printable ASCII?

[Tufty]

I might be missing something, but I thought we had to find both an Initial Vector as well as a key.

Does this new code automatically assume an initial vector while rotating through the keys?

[c1023]

[BrianEnigma]

[Langley Moor]

Am I right in thinking if it comes back with a non-sensical answer it doesn't say anything? Am just running a few ideas through it!

EDIT: Right, am currently running 100,000,000 passwords starting from aaaaaaaa, just to see if I can catch something. If someone wants to start at a later point and keep going, we might just strike lucky? I assume this thing is only putting in printable characters as possible passwords?