On 06/10/2017 8:31 AM, James A. Donald wrote:
Generating this proof is inconveniently slow, and the data that has to be stored is inconveniently on the fat side, but verifying it is reasonably fast.
I believe the issue with this, such as in Bitcoin and any coin, is that the blockchain eventually grows to an enormous size. It's all nice when it's at around 200GB of data and it only takes a few days to synchronize to the network. But think about a blockchain that is 2 TB. That is not only a huge blockchain, that would be huge for a relational DB that is stored on central server or cluster. When databases get to that size, companies divide the schema. They'll either divide it per system, per department, per branch or however it suits them best. Then they'll use some sort of business intelligence tool to make sense of all that data. Nodes need to be able to be useful by storing smaller chunks of a blockchain instead of having to store the entire thing, something like how sharing a bittorrent file works. The downside is that this may lead into centralization as full nodes would have an advantage, but it's not worse off than what it is now. ./gv