Justus Ranvier, crypto commentator and reviewer with Open Bitcoin Privacy Project, proposed in an essay that the idea of block size limits is based on economic fallacy. He writes that if price discovery were instead introduced to the Bitcoin network, no block size limits would be needed.
“My proposal is: stop trying to solve the problem of how to make sure people don't use too much of a resource that's being provided for free. Instead, figure out how to allow users of network resources to pay the providers of those resources.”
— Justus Ranvier
In terms of application, once the “price” of any given transaction is discovered, says Ranvier, Bitcoin's nodes, miners, and users could pay one another directly for bandwidth, information and hardware-driven services via existing micropayment channels.
It seems almost political: The Bitcoin block size debate is presented as having only two sides – like red vs. blue. Republican vs. Democrat. Conservative vs. liberal.
And as is common in politics, this two-options-only paradigm may also be false when applied to Bitcoin. We asked Ranvier to tell us more about his idea for a “third option.”
“Any attempt to set a protocol-defined policy relating to the block size would require the designers of the policy to solve the economic calculation problem in order to create the ideal policy. Since that's not possible, any policy chosen will result in misallocation (shortages).”
Cointelegraph: You argue that both the "do nothing" approach and the 20 MB cap idea are missing the point. What is the point, exactly?
Justus Ranvier: The point is that nobody knows, or can know, the correct number of transactions the Bitcoin network should process, or at what price they should do so.
Any attempt to set a protocol-defined policy relating to the block size would require the designers of the policy to solve the economic calculation problem in order to create the ideal policy. Since that's not possible, any policy chosen will result in misallocation (shortages).
CT: How would price discovery among users, miners and nodes work?
JR: Right now, the Bitcoin network has a problem with bandwidth allocation due to the fact that every full node gives away as much bandwidth as other nodes ask for. It means that as the number of users increase, the financial burden on node operators who are donating their resources also increases.
As in all "tragedy of the commons" problems, the solution is to abandon collectivization and allow a market to form that connects producers and customers.
My proposal is: stop trying to solve the problem of how to make sure people don't use too much of a resource that's being provided for free. Instead, figure out how to allow users of network resources to pay the providers of those resources.
In that scenario, each node in the network acts as an independent agent who attempts to obtain information other entities want to purchase, and then sell that information to them. For example:
- End users want their transactions to reach miners.
- End users want to receive information about their wallet balance.
- Miners want to receive fee-carrying transactions to include in blocks.
- Miners want other miners to receive their blocks, and want to receive the newest block.
If full nodes can trade with both miners and end users to route transaction and block information, then competition between them will result in price discovery for network bandwidth.
If the network bandwidth is paid for by the users of the network, rather than donated by node operators, then we no longer have to worry about the network consuming "too much" bandwidth.
CT: How would your proposal incentivize further decentralization in Bitcoin?
JR: I don't know if it would or not, because I don't know what you (or anyone) means by "decentralization."
What I can say is that market allocation of services is the best way to satisfy the combined desires of producers and customers, in any context.
CT: Which technologies in your proposal already exist, and which still need to be developed?
JR: All of the individual technologies needed to create the market I described exist, but they've never been integrated in this particular manner before.
CT: Which of them only need wallet integration, and which need actual Bitcoin protocol implementation?
JR: If by the "Bitcoin protocol" you mean the consensus rules, then the only change needed is to remove the block size limit once it is no longer necessary.
The actual changes needed to create price discovery between nodes would be confined to the network communication protocol. Wallets and nodes would both need to implement support for some new communication protocol features.
See Ranvier's “Economic Fallacies and the Blocksize Limit” for more technical detail and answers to frequently asked questions.