Something that hasn't been done much yet, but probably will as the space matures, is for a blockchain to reuse another blockchain as its "timestamp" generator.

Something that hasn't been done much yet, but probably will as the space matures, is for a blockchain to reuse another blockchain as its "timestamp" generator. eg. A 'MyCoin' miner grabs the latest 10 blocks off the main, say, Ethereum blockchain, and includes those hashes as proof they began their block mining _after_ those main blocks were mined, then adds their own proof-of-work on top, winning the block-solve lottery to add the next block to the blockchain. This could potentially reduce the amount of relative advantage a dominant miner has ( others can provide a math analysis ) This approach can also save electricity, as it puts the miners on a mutual 'duty-cycle' where MyCoin miners are only busy running cryptohashes for a fraction of the time between blocks. Another nice side-effect of this 2-tier mining is that the block generation times are spaced more evenly over time. [ They move towards a normal distribution around the average time, rather than uniformly - a result of the 'law of large number's, which states roughly that the average of uniform 'flat' distributions approaches a normal bell curve around the mean ]. Currently Bitcoin blocks sometimes arrive after 1 minute, other times only after 19 minutes, with mean 10mins maintained by adjusting difficulty. So the bitcoin time between blocks variance is high ( Litecoin adds blocks 4x as often @ 2.5mins, Ethereum ~10secs on average, but all with quite high variance ) which can surprise users.