”Closing” the blockchain

In accounting, you close the yearly ledger and create a new to which you transfer some of the values from the prevoious year.

I know this as a programmer. The company I worked for had one table for every year.

I think it s something that you could and should do with a blockchain because of how fast it grows. Before long, the blockchain will be humongous

So, at some point it should be possible to say that the cuurent blockchain van be closed and archived. A new genesis block or blocks need to be created and they should contain the balance of each address where there are coins.

There was a locked article in a magazine that mentioned that some people had come up with an idea for a blockchain that could do this. It seems like a very reasonable thing to do. Sooner or later the sheer size of some blockchains will be a very real problem.

Compressing settled state combined with a way to distribute only the most recent ledger history via relay nodes while keeping the full chain on block producing nodes only may be good enough to achieve decentralization with more scalability. This puts the burden of maintaining the complete history of the ledger on SPO but ideally they also have the most incentive to participate fully in the network.

The size of the blockchain will be measured in petabytes soon. The higher the tps, the faster it will grow. It seems solana is already there.

Solana works differently, the size is partially due to adoption but also the tendency to use micro-transactions to effectively complete programs in multiple steps. Closing out the ledger at a given time period using a hash that then gets folded into proof of history from their perspective makes a lot more sense.

Over here on Cardano I suspect they will always need at least some nodes that have a full copy of the blockchain for security and decentralization. Cardano favors larger and less frequent transactions but not just because the TPS is low. Compressing and/or partitioning ledger state would make more sense here for performance.