The biggest bolt-on service to the bitcoin platform these days is mining pools. The success of mining pools illustrates both a triumph and a flaw in the bitcoin economic theory. The sheer number of individual miners is fantastic for the stability of the system, but the existence of centralized pool operators is bad for system security and illustrates the flaw in the system. Difficulty has simply become too high for individual miners to play with. Mining is only worthwhile with a pool.

If this is addressed in bitcoin itself, the system reverts to its usual secure, p2p nature, and becomes even stronger. The success of the mining pools themselves proves that this is a crucial feature. One option is to bake pooling into bitcoin. A miner can either do conventional mining, but at much lower difficulty as he would today in a pool, or he can fulfill the role of a pool operator and aggregate blocks. This would require some clever tricks to make sure miners and operators don’t cheat each other, but the payoff is huge.

Another option is to keep difficulty at a flat, low level, and instead of raising difficulty, split the block chain into ever more branches. Each miner would work on the block chain that corresponds with his bitcoin account (via a modulus operation), and work only on transactions that correspond with his bitcoin account. This has the added advantage of balancing work among many nodes. The major issues here are how to prevent double spending, since older transactions could be in any branch, and whether the security of the system would be compromised, since it would be much easier to attack any one branch.

I'm sure there are many other ways of tackling this, but the point is that mining pools are a clear signal pointing out perhaps the biggest problem yet discovered with bitcoin.