DarkBlizz

I just thought I'd post, the installers (bootstrapped) for enUS, enGB and zhTW are all identical.  This leads me to believe that there still may by separate keys for/as each version points to it's on web address for the authentication key.

If anyone has source for bruteforcing*, I have over two teraflops at my disposal.  The standard 2^256 would still take a significant amount of time, but a reduced set may be within the realm of possibility.

Actually, limiting the character set to 32 allows them to express a 2^5 sequence as a unique character with no leftovers. 

It's odd, 32 char set with a 32 char key... this is well above the 256-bit salsa key (4x).  It just seems to me that Blizzard could have this with irony...


Anyways, somebody want's to drop me some code, I can run it. 
Kremnari


P.S, has anybody checked for differences in the enGB vs enUS installers?

*Either direct bruteforcing the mpqe, or generating a list of potentially valid keys would work. 

I've been paying attention to this since the first thread opened up, and here's something I'm not getting.  The authentication key is supposedly 30 characters (assuming blizz didn't add useless characters to the string); is alphanumeric (sans 0,o,1, and i; leaving a 32 character set); salsa20 uses a 256-bit key (again assuming standard implementation).  This means that the mpqe's should only be (truly) protected by an 8 character code (from their 32-bit character set).

From what I'm reading, the authentication code should either go through some kind of conversion to become the mpqe key, or the mpqe key should be stored somewhere.  It seems that most people get this.

Hopefully atleast one person who can follow assembly is working on this, so my question is:  What does the installer do when you manually type in the code?  A 30 character string can't just disappear.


I'll try looking at it again myself, see if inspiration hits.
Kremnari