The comment below proposes a way to get all the purported benefits of both small and large blocksize limits without the purported downsides of making either choice:
This is in several ways more attractive than the fork arbitrage idea I've been pounding the table about, because it allows people to use both systems exactly when and exactly for as long as they are both useful, without messing with hodlers at all (crucial!).
Besides a few possible reservations regarding the timing of spinoff ledger-snapshots, this model looks to enable something no one seems to think possible now because of failing to fully internalize the fact that Bitcoin is economically speaking just a ledger and set of private-public keypairs controlling those ledger entries (along with the coin issuance schedule). Also that two ledgers are functionally the same as just one ledger, in a way that the "split" can be abstracted away from the user experience. And finally that preserving the investments of people who already bought into the World Wide Ledger is the only thing that's necessary to maintain Bitcoin's monetary network effect (new investors would have to choose their ratio of coins to buy in the two forks, but present investors lose no upside from that; the investor promise is as it has always been).
Most people understand none of these three points, so they miss what I think is a beautiful solution if it works technically. Indeed the debate would be over, and perhaps almost all similar debates would never need to happen other than as debates about wise investment decisions. This seems like the ultimate expression of "Bitcoin is a creature of the market." Not only would multiple implementations prevent every node from failing at once, it would prevent even any price loss when it happens! (unless you weighted your portfolio to favor the failing implementation)
This means Bitcoin can effectively have both a tiny cap and a huge cap (or no cap), because Bitcoin is the ledger and its monetary parameters, not the protocol. The motto that kicked off the drive for Bitcoin Unlimited was, "The blocksize cap is an aspect of the transport layer that was falsely considered to be in the consensus layer." Or, "The blocksize cap snuck into the consensus layer when it should have been in the transport layer." Similarly perhaps we say something like, "The non-monetary aspects of the protocol are part of the ledger maintenance function that were falsely considered to be inextricably tied to the ledger itself." Or, "The non-monetary aspects of the protocol snuck into our conception of what Bitcoin is, when they should have been just part of how the ledger is serviced."
In other words, different implementations of the protocol are just different competing service layers built on top of the basic ledger layer. Again, since two ledgers are in principle no different than one (a collection of balances tied to keypairs along with another collection of balances tied to [EDIT: those same] keypairs is still just a collection of balances tied to keypairs), diverging ledger forks do not functionally imply multiple ledgers from the perspective of investors who have already bought in at any given time, which is all that matters.
Functionally, from the point of view of every already-invested user at any given time, there remains one universal ledger of civilization, no matter how many protocols service it. Wouldn't it be neat if Core were like Breadwallet, just another service operating on top of Bitcoin? That's the implication here, if I'm not mistaken.
Also, although the reddit comment says merchants would need to demand payment in both coins at once in order remain neutral (investment-wise), which seems a hassle because you'd have to wait for confirmations in both forks, this isn't strictly true: insofar as there is a stable market price ratio between the two, merchants can accept payment in whatever is most convenient and cheap (likely the big block fork, but who knows) and rebalance their portfolio periodically if they chose to.
Some issues with spinoffs:
1) The spinoff timing is crucial because there has to be a clean Schelling point for when to take a snapshot of the ledger. For example if there are several spinoffs that are all the same except their snapshot timing differs by a few blocks, how would people choose one? There would have to be a big advertising push. Or would there? Perhaps there is a way the market would sort it out that I'm not seeing yet.
2) Related the previous point, a spinoff would likely not happen until there was enough of a "crisis" situation to fuel a sufficiently clean Schelling consensus and also entice plenty of investors to arbitrage the two ledgers against one another.
3) There are probably some disruptive implications for the mining industry in the event the split becomes permanent. For example, if the spinoff uses a different mining algo and the coin price ratios end up at 50/50, the current crop of miners would instantly lose half their profitability and there would be a race to develop ASICs for the new algo. He halved hashrate wouldn't lower security for Bitcoin, though, because each set of mining infrastructure would only be securing half the total universal ledger's value. This may mean spinoffs are a scaling solution all by themselves, since you can have as many spinoffs (i.e., ledger-servicing clients) as you want. Alternatively, if the spinoff uses same mining algo, I'm not sure how 51% attack risk might be affected. However, if one spinoff is 51% attacked, neutral investors lose nothing so it actually makes the system more robust from an investor's standpoint. You have to kill ALL the ledger-servicing clients to have a chance of harming the ledger or the price.
This is in several ways more attractive than the fork arbitrage idea I've been pounding the table about, because it allows people to use both systems exactly when and exactly for as long as they are both useful, without messing with hodlers at all (crucial!).
Besides a few possible reservations regarding the timing of spinoff ledger-snapshots, this model looks to enable something no one seems to think possible now because of failing to fully internalize the fact that Bitcoin is economically speaking just a ledger and set of private-public keypairs controlling those ledger entries (along with the coin issuance schedule). Also that two ledgers are functionally the same as just one ledger, in a way that the "split" can be abstracted away from the user experience. And finally that preserving the investments of people who already bought into the World Wide Ledger is the only thing that's necessary to maintain Bitcoin's monetary network effect (new investors would have to choose their ratio of coins to buy in the two forks, but present investors lose no upside from that; the investor promise is as it has always been).
Most people understand none of these three points, so they miss what I think is a beautiful solution if it works technically. Indeed the debate would be over, and perhaps almost all similar debates would never need to happen other than as debates about wise investment decisions. This seems like the ultimate expression of "Bitcoin is a creature of the market." Not only would multiple implementations prevent every node from failing at once, it would prevent even any price loss when it happens! (unless you weighted your portfolio to favor the failing implementation)
This means Bitcoin can effectively have both a tiny cap and a huge cap (or no cap), because Bitcoin is the ledger and its monetary parameters, not the protocol. The motto that kicked off the drive for Bitcoin Unlimited was, "The blocksize cap is an aspect of the transport layer that was falsely considered to be in the consensus layer." Or, "The blocksize cap snuck into the consensus layer when it should have been in the transport layer." Similarly perhaps we say something like, "The non-monetary aspects of the protocol are part of the ledger maintenance function that were falsely considered to be inextricably tied to the ledger itself." Or, "The non-monetary aspects of the protocol snuck into our conception of what Bitcoin is, when they should have been just part of how the ledger is serviced."
In other words, different implementations of the protocol are just different competing service layers built on top of the basic ledger layer. Again, since two ledgers are in principle no different than one (a collection of balances tied to keypairs along with another collection of balances tied to [EDIT: those same] keypairs is still just a collection of balances tied to keypairs), diverging ledger forks do not functionally imply multiple ledgers from the perspective of investors who have already bought in at any given time, which is all that matters.
Functionally, from the point of view of every already-invested user at any given time, there remains one universal ledger of civilization, no matter how many protocols service it. Wouldn't it be neat if Core were like Breadwallet, just another service operating on top of Bitcoin? That's the implication here, if I'm not mistaken.
Also, although the reddit comment says merchants would need to demand payment in both coins at once in order remain neutral (investment-wise), which seems a hassle because you'd have to wait for confirmations in both forks, this isn't strictly true: insofar as there is a stable market price ratio between the two, merchants can accept payment in whatever is most convenient and cheap (likely the big block fork, but who knows) and rebalance their portfolio periodically if they chose to.
Some issues with spinoffs:
1) The spinoff timing is crucial because there has to be a clean Schelling point for when to take a snapshot of the ledger. For example if there are several spinoffs that are all the same except their snapshot timing differs by a few blocks, how would people choose one? There would have to be a big advertising push. Or would there? Perhaps there is a way the market would sort it out that I'm not seeing yet.
2) Related the previous point, a spinoff would likely not happen until there was enough of a "crisis" situation to fuel a sufficiently clean Schelling consensus and also entice plenty of investors to arbitrage the two ledgers against one another.
3) There are probably some disruptive implications for the mining industry in the event the split becomes permanent. For example, if the spinoff uses a different mining algo and the coin price ratios end up at 50/50, the current crop of miners would instantly lose half their profitability and there would be a race to develop ASICs for the new algo. He halved hashrate wouldn't lower security for Bitcoin, though, because each set of mining infrastructure would only be securing half the total universal ledger's value. This may mean spinoffs are a scaling solution all by themselves, since you can have as many spinoffs (i.e., ledger-servicing clients) as you want. Alternatively, if the spinoff uses same mining algo, I'm not sure how 51% attack risk might be affected. However, if one spinoff is 51% attacked, neutral investors lose nothing so it actually makes the system more robust from an investor's standpoint. You have to kill ALL the ledger-servicing clients to have a chance of harming the ledger or the price.
Last edited:
