Quote from: matt.collier on May 22, 2011, 03:28:33 PM

Are we using a 160 bit hash (which provides for the possibility of a collision) vs a 224/256 bit hash (no possibility of a collision) so that bitcoin addresses can be shorter in length? If so, is it possible for us to transition to using a 256 bit hash at some later date?

I don't buy the argument that it's TOO computationally expensive to intentionally create a collision. We have already seen the use of GPUs radically alter the bitcoin mining paradigm. In the future, we may well see devices designed specifically for the task of performing hashing functions. Perhaps those devices already exist.

Why build the opportunity for fraud into bitcoin? I don't think we need to be concerned about the number of characters of a bitcoin address when we're copying and pasting them or using QR codes anway.

Don't worry, "If you were to intentionally try to make a collision, it would currently take 2^126 times longer to generate a colliding bitcoin address than to generate a block.". This means, if you have a computer that is 1 million times as powerfull as all current miners combined, it will still take an average of 1,618,542,460,620,902,128,345,579,373 years to generate a collision.

Even if Moores law holds true in the most generous way, we still have over 100 years left before this becomes feasable.

And also: yes, devices designed specifically for performing bitcoin mining exist (Artforz had himself some ASICs (custom chips) made)