Re: Bitcoin XT - Officially #REKT (also goes for BIP101 fraud)
How can we loose that free choice if anyone is free to release an alternate client the same way XT would have. If it get enough support it will simply be adopted.
Yes, at this point, the block size limit debate is sound and fury, signifying nothing. If the market wants to increase the block size limit, then it will increase the block size limit.
I've shown these diagrams a lot because I think they reveal the essence of the situation. If the limit remains to the right of Q*, then it doesn't really matter what the limit is because it does not affect the free market dynamics. However, if the limit falls to the left of Q*, then the pressure due to the deadweight loss will eventually cause a fork to move the limit back to the right of Q*!
TL/DR: There is no way to stop Bitcoin from growing.
Peter, your economic analysis is flawed, as it simply wishes away the external costs that the market may choose to impose on the users running their own node. The scaling solution needs to take account of this, and your proposed solutions as well as your analysis does not. Stop promoting faulty ideas (economic and technical alike).
What Peter propose makes a lot of sense economically as the market will want that dead loss regardless of the nodes but I get your point that it does not take security in consideration.
Are you suggesting the anti spam measure should take into consideration the number of nodes in the network? It might make sense to make a rule that enforce the market to have a minimum number of nodes in order to let the blocksize grow. The market will have an economic incentive to have a minimum amount of nodes in order to get that dead loss.
However, what's the minimum of node should realistically be enough to consider bitcoin secure enough? How much blocksize growth should it let if the conditions are met? How do you get the balance between the acceptable amount of dead loss VS the acceptable amount of nodes?
The number of nodes is irrelevant.
The block size limit should be a measure of the cost of the option to create a full node
Cost is relative. It might be cheap for you but not for me. I also don't think it would be possible to enforce it into the protocol.
"Nigga what?"
Cost is cost. Let's call it a price if you fancy it doesn't matter.
Of course it is possible. Satoshi did it for the monetary supply, the block interval and the space in blocks. All of these are enforced today by what of the Nakamoto consensus protocol. None of these rules will change unless the economic majority agrees to it.