Subject: Re: rpc xid randomness
To: None <>
From: Jun-ichiro itojun Hagino <>
List: tech-security
Date: 09/09/2003 06:10:44
> It looks as though the value is calculated from:
> 	a ** b mod c
> where 'a' changes for each block of numbers, 'b' sequences through
> terms of a LCG whenever a value is wanted (missing 0 to 7 each time)
> and 'c' is constant.
> Some notes I have on the security of RSA (where 'a' would be the message
> and 'b' and 'c' the key) say that you should not use different values of
> 'b' with the same 'c' - otherwise recovering the key is trivial if the
> same message (ie 'a') is encypted with both.
> This use of the equation seems to be going out of its way to make it easy
> to break!

	we're not encrypting stuff, we're just trying to generate non-repeating
	random number.  so i beg to differ.