Even simpler minimum fee calculation formula: f > bounty*fork_rate/average_blocksize [combined summary]



Individual post summaries: Click here to read the original discussion on the bitcoin-dev mailing list

Published on: 2013-11-20T10:01:08+00:00


Summary:

In an email exchange on November 15th, 2013, Peter Todd and Michael Gronager discussed the speed of Bitcoin in terms of alpha. Todd calculated alpha to be 39.62 microseconds per byte or 24 kilobytes per second, which he considered to be inefficient. Gronager then calculated alpha to be 29 milliseconds per kilobyte using a different formula. They debated where Gronager got the number 454,000 from, and Gronager explained that he used the actual block interval instead of the steady state one. The conversation ended with Todd's signature.The primary focus of the conversation was the efficiency of block propagation in Bitcoin. Todd and Gronager calculated the alpha value for Bitcoin to be 24kB/s for block propagation, which they deemed inefficient. Gronager suggested that optimizing profits could be achieved by using high-bandwidth nodes with restricted numbers of peers, as mining pulls in $1.8 million per day and not optimizing profits results in a loss of up to $16k. However, smaller miners bear this loss disproportionately. They also discussed the advantages of P2Pool in block propagation efficiency but acknowledged the challenges in growing P2Pool due to the requirement of running a full node.Another topic of discussion was the economics of mining pools in Bitcoin. Gronager argued that participating in a pool can be advantageous due to economies of scale, despite miners having to pay a fraction of their income to the pool owner. Todd agreed that there is an incentive for miners to join pools and centralize, but regulation or social pressure could counter this. They delved into equations and models to analyze the profitability of mining in pools versus mining solo, considering factors like hashing power and block size.The email thread also covered topics such as transaction fees, orphaned blocks, and the impact of mining centralization. They discussed the minimum fee for Bitcoin transactions, the relationship between block size and latency, and the potential implications of relaying all orphaned blocks. They highlighted the need for an optimal tx inclusion policy in P2Pool and emphasized the importance of network and node latency in the scalability of Bitcoin.In summary, the email exchange provided insights into various aspects of Bitcoin mining, including block propagation efficiency, the economics of mining pools, and the implications of mining centralization. The conversation involved calculations, equations, and analyses to understand the dynamics of the cryptocurrency ecosystem.Blockchain.info's sybil attack on the network revealed that the average fee for bitcoin transactions over the last 90 days is approximately 0.0003BTC/txn, close to the theoretical minimum of 0.00037BTC/txn. However, this may be due to old clients and fee settings rather than network wisdom. Gronager conducted calculations using different pool sizes and found that the measured fork rate can estimate the minimum fee. He determined alpha as a function of the average fork rate and average block size, which yielded a minimum fee formula of f > P_fork*E_bounty/S_average. By analyzing historical data from blockchain.info, Gronager determined that the minimum fee should be f > 0.00165XBT/kb or 0.00037XBT/txn based on the average number of orphaned blocks per day and the average block size over the past year. Looking at the trend over the last 12 months, Gronager found that considering only the last 90 days, the minimum fee would be f > 0.00162XBT/kb or 0.00037XBT/txn. He also noted that a 4 times increase in block size would still be sufficient for the current load based on the optimal revenue block size being a function of the transaction fee. The key factor in determining fees is alpha, which represents network latency and depends on interconnectivity, bandwidths, and software quality. The email exchange concluded with the understanding that software quality improvements could bring down fees, as standard clients can be configured to choose a better fee based on easily measured parameters.


Updated on: 2023-08-01T06:32:56.391720+00:00