 |
August 09, 2017, 05:08:38 AM |
|
I'm interested in creating a fork of bitcoin that improves its scalability in a fundamental way. Using a volume model, each individual miner has fewer transactions to listen to and store.
The issue is of course programming. I've got a fork of the code and I've begun working on the protocol, but it will take a very long time to complete I suspect. I'm hoping this thread will serve as a way to learn some important things about bitcoin's implementation as well as find some people to help design tests for the network once I've got a prototype.
My first question is whether block rewards could be generated by transactions directly, as in a block with 150 kb of transactions generates 150 uBTC before transaction fees are applied, while a block with 500 kb of transactions generates 500 uBTC. What are the possible abuses of this system, and how can they be mitigated? My though was that the minimum fee would be twice the transaction reward, and the transaction reward would go down as the currency matures (to avoid runaway inflation). Other than that, what should be done?
Secondly, what issues does a fast block time cause? (10s per block, for instance)
Third, why is txn_count in block headers always 0 in bitcoin's implementation? This currency requires transactions (or total transaction data more specifically) to be tallied in every block header for a few purposes. Is there a problem with doing this?
Thanks in advance for your input! I'm very excited to make this project a reality, because I think this currency could provide scalability and longevity in a robust way.
|
