CheckRevision (BNLS users) questions

[warz]

Users of the new BNLS, I come to you with a question.

If you tell bnls to use the same lockdown library #, and provide to it the same "hash value" from SID_AUTH_INFO, will it produce the same checksum value over multiple calls to checkrevision? Or will the checksum values change from call to call (keeping all arguments to it constant).

Please, be on with this test, good peopl! And let me know thy results!

[Hdx]

Well obviously it would result in the same value.. Or else Blizzard would have no way of telling if  you got the correct value....
~-~(HDX)~-~

[Kp]

Not necessarily.  They could cause the results to vary over multiple runs by incorporating factors such as the IP address of the BNCS or the time of the connection (with some bits discarded to deal with clock drift).  Any quantity which can be measured by both parties is a candidate for inclusion in the result

[warz]

Well, I wasn't really asking for theory, considering we can test this by using bnls. I have no client though that uses bnls, much less the source code to one. There's no reason to guess about this.

[Kp]

You can make one in ~5 minutes.  I did last night. :p

Testing against BNLS doesn't really prove anything though.  Even assuming Skywing is providing results that exactly match what lockdown does, there remains the possibility of a time-based component.  How large a set of times all produce the same result?  That is, do I have to wait a minute, an hour, a day, or a week before lockdown's time sensitivity changes the result it gives for otherwise-static input?

[Ersan]

Told you in IRC, but here for anyone else that wants it:

Seed: ÐíŸìRÛÊÑz?iX3Z_
mpqName: lockdown-IX86-05.mpq

Attempt 1:
(long) Result: 2050155150
Attempt 2:
(long) Result: 2050155150

tried it a bunch more times, on different days, still the same.
(that's a yes)