RE: The Concept of Hierarchical Self-Purging Blockchains Explained
You don't delete the blocks. Let's say there is an exchange or merchant that does 10,000 Tx a day. We can use 3 second blocks to do the Tx and each day there is another blockchain that record the net effect of the exchange/merchant as a single Tx. On the next day, the 3 second Blockchain can reset itself and get the data from the 24 hour blockchain.
Think you put some Bitcoin into Binance and do 25 different trades in 5 minutes and converted everything back into BTC and transfer it to your BTC wallet. The blockchain is only going to register your 2 Tx with Binance. There will be no records of the other 25 Tx on BTC. It'll be only on Binance servers. My proposal is on chain solution where multiple interoperable blockchains with different blockchains. The higher level blockchains will record the net effect of lower level blockchains. After this net effect is recorded, you can restart the lower level blockchain using the higher level blockchain as a reference.
OK. This is just a form of side chain. You put a starting transaction in the top level chain, run a series of transaction in the lower levels and accumulate them. When you're done then you make a top level ending transaction with the accumulated activity .
Isn't this essentially how the Lighting Network works?
Comparing LN and my concept is like comparing Ethereum and EOS. They are kind of going for the same hing but there are fundamental differences. My concept is 100% on-chain. It slows down the growth of the blockchain size greatly. There are also added features like selecting your Tx speed/cost/risk. It's also easier to deal with attacks. Komodo uses delayed POW (they recycle the BTC mining power essentially making it the most secure proof of work platform). Each level of the blockchain concept I discussed can be a different blockchain with its own security. If somebody breaks SHA-256 BTC is done. Dash actually has 11 algorithms chained together. But with my proposed system, you could use practically anything and everything.
The only major downside to what I propose is it'll be a total hell to develop. A smart ontract platform is just for other people to run DAPPs on it. But what I propose is to make a mesh of blockchains. The difference is like setting up a simple LAN at home Vs a decentralized P2P global network. It's going from http to IPFS.
Do you have a formal description or a white paper? I'm working on a mesh concept of my own. A true mesh allows for things block chains don't like - things like cycles in the graph.
No. I'm not a programmer or anything. I've only been into crypto since last year. I was simply trying to get rid of the need for a total history to keep the blockchain immutable.