KeYgenme.7.0.ECrypt.Challenge.2008.FOFF.

[Encrypto]

well well well ...

a lot of time spent on this one

A WORD OR TWO OF WARNING :

this challenge is not intended for complete n00bs. partly because

serialphising is impossible unless your really clever

[Encrypto]

no one ?

[Fellow Hacker]

Give Us Time Encrypto We All Aren't As Smart As You.
Laterz,
Short-Circuit

[Fellow Hacker]

Ok Well I Suck With Keygenning So All I Can Come Up With Is That 1F Is Used A Bit In ECX And CDQ (Convert Double Word To Quadword) And Then There Is Integer Division Involving Use Of The Name And That Integer Division I Mentioned Earlier?
It Is Used With Hard-Coded Letters And Numbers.
Hope I Got It Right But I Suck With Keygenning So I Might Be Wrong.
This Was Just My General Look On This KeygenMe
Laterz,
Short-Circuit

[Encrypto]

There are no hardcoded strings in there

well .. other than some which are used for simple Ciphers.
then the cipher result is passed on to a HEAVILY modified md5.. which
according to some people they might say that its not md5.. but i did
base it on md5 and i call it ECrypt.

Then there is RSA involved which is also quite hard to understand..
i am still gripping its basics lol .

you are partly right FellowHacker. IDIV in asm is the mod function in programming. In C++ i think its the % function. and in delphi its just mod.
the result from the mod is then used as a pointer to a table which gives a result

a bit too much to give away but oh well

Im not smart FellowHacker. Its a LOT of help from my good friend
SerialKiller. Also Mynes for the superb gFX

[chupachu]

I just got back from my holidays.. On a quick look its RSA, i will look at it in next days if i got time, i havent done much rsa actualy, so it will be a good thing to learn stuff for me :)

[Encrypto]

Hehe. Welcome back chupachu

some special surprises in there

Have fun

[chupachu]

If it really based on md5 then you have really done some HEAVY
modifications to it, if it will require too much of my time i will surely
not keygen this to end.

I just got to init part of moded md5, but i got a feeling figuring out whats
all moded beside 2 aditional magic values and new transform variables
will be PITA :)

Good keygenme this time, its zilion percent different then all others you
have coded, as if you woke up with new brain the day you coded this ;)

BR, ChupaChu!

[Encrypto]

Indeed i did wake up one day .. and this idea poped into my head.

:P

SerialKiller helped me figure out what was going wrong and how to fix it

thanks for the comments Chupachu .

Much appreciated.

[the_one]

Hi Guys!


im on a borrrowed computer so i don't ahve any of my crackjing software here ):

but it looks cool so i will sertanly attempt it when i get my own back
but i really doubt i can make a keygen :S

[chupachu]

I lack RSA experience to be absouletely certain, but i had a lot of
headache to adjust all md5 modifications, and later on when i tought
tha i had keygened it successfully i found that it does not work 100%.

It work for most of normal words, and names, but when using
random strings (usually longer ones) it will produce wrong keys.

I am certain it has to do with md5 modded part but i cannot find
where the problematic part is. And due to lack of free time i cannot
perssue it anymore.

After sending my keygen to Encrypto and recieving full sources of
ECrypt i thought i coud figure out where i miscoded my md5
modifications, but it seems i cannot find it.

IMHO problem could be in variable type LongWord that ECrypt uses
its 0 to 4294967295 and in calculations it always is positive (>0)
and in normal md5 its DWORD its also in same range, but it can be
interpreted as negative as well so in functions as II, FF etc one actually
expects additions, but md5 performs subtractions aswell, so my guess
is somewhere along the way problem accours, one can call it a bug,
other a feature ;)

I guess that's the beauty and ugliness of modifying cryptos one did not
invent, it can come out good or bad, depends on the way you look at it :)

Another thing is that when message (e.g. name) is longer than
256 bits e.g. (M) > (N) what shud mean no correct decyphering
would be possible even with a correcIt private key (D).

If someone wants to figure this semi-mistry to the end, PM'me and
i will send you mine sources. And if Encrypto agrees i can send his
original ECrypt unit also (to be easier to compare).

BR, ChupaChu!

[Encrypto]

sure why not

this is only a challenge ..

XD

[hmx0101]

Mate take a look to my srcs too
I've keygenned it, some weeks ago

[Encrypto]

Well the md5 digest wont me more than 256 lmao

how can it ? its core function produces a 128 bit hash or 32 length.

[chupachu]

Take look at original, your and mine version of md5 mods.