Re: stamp collisions

> With v1 stamps the full collision string is used for avoiding stamp
> collisions because there is a separate field for counting to find
> partial hash collisions.  With the C implementation this defaults to
> 96-bits or 2^48 stamps before 1st collision, which is pretty small.

The RNG of some implementations is prone to occasionally working from 
the same seed at the same (or similar) time.  In other words, if Bob 
and Charles both send a mail from their Windows machines (or some other 
platform without a strong RNG facility) to Alice at the same time, they 
might inadvertently produce the same random field, with a far greater 
probability than a true randomness model predicts.  Without extensions, 
this will often result in identical stamps.

This is despite the precautions taken in the code to introduce entropy 
to the seed from several different sources.  Because none of those 
sources are truly random, the resulting seed is not either.  Strong 
RNGs typically take entropy from sources of "environmental noise" such 
as mouse movements and interrupt timings, which carry at least small 
amounts of true randomness.

To eliminate collisions entirely, a "sender=..." extension field would 
be advisable.  This reduces the chance of collisions to 1 in 2^48 
stamps produced by any particular sender, even if they only have a 
medium-quality RNG.

--------------------------------------------------------------
from:     Jonathan "Chromatix" Morton
mail:     [email protected]
website:  http://www.chromatix.uklinux.net/
tagline:  The key to knowledge is not to rely on people to teach you it.