[Cryptography] BitCoin Question - This may not be the best place to ask

Steve Weis steveweis at gmail.com
Sun Dec 22 20:59:36 EST 2013


On Sun, Dec 22, 2013 at 4:30 PM, Robert Christian
<robertjchristian at gmail.com> wrote:
> 2) I am pointing out that addresses are finite, and 34 chars long... They
> can only be upper or lower case, or 0..9.  So at the end of the day, after
> all the fancy stuff, the number of all possible bitcoin addresses is
> (26*2+10)^34 possible unique ids.
>
> So the number of possible unique addresses is actually relatively smalll.
> Right?

The address has 20-bytes of hash, a network ID byte prefix, and a
4-byte checksum. So, there are 2^160 possible unique addresses. This
is converted into a 34 character base-58 string.

You do bring up one point that many key pairs will collide for a
particular address. That's why the hash function must be assumed to be
collision resistant.

As for when we might see collisions, with a birthday attack you'd
expect there to be a 50% chance of some collision existing when there
are roughly 2^80 addresses.


More information about the cryptography mailing list