The possible complication is if the community gets stuck in a half-way state, with almost equal mining power on each side. Then there may be two blockchains, lots of orphaned blocks and branches, etc..
It was my understanding the XT fork will only take place if 75% of hashpower supports it. I suppose miners could change their minds after the fact, but otherwise a "half-way state" will not result in a fork.
The BitcoinXT version may never conquer 75% fo the hashpower. Then there will be no fork, and people can continue using either version, indifferently, modulo future upgrades and forks.
Suppose the BitcoinXT does get 75% of the hashpower. Since BitcoinXT implements a grace period between the 75% vote and the actual lifting of the limit, there will be time to alert all remaining Core players to upgrade, and hopefully they will do so. Then only the XT branch will survive, and there will be no other complication for the clients.
However, if the Core players are stubborn, the two branches will persist for a while. There would be
a "XT" branch, with more PoW, mined by the XT miners and accepted by all XT players and (IIRC) by the SPV Core clients; but also
a "Core" branch, with less PoW. mined by the Core miners and accepted by full Core nodes only.
If the ratio were to remain 75% XT : 25% Core, thenThe XT branch may start out with one block every ~13 minutes, and the Core branch with one block every ~40 minutes; but eventually their difficulties would readjust so that both will have 1 block every 10 minutes.
Given the way that the way the fork is programmed, every transaction issued will be executed in principle on both chains; so, for most clients, there will seem to be only one coin. However, by accident or intentionally, a transaction may eventually include fresh coins mined after the fork, and therefore will be valid only on the branch where those coins were mined. An expert bitcoiner could exploit that trick to sell one version of coins and buy more of the other, or spend each version independently. However, as soon as the impasse resolves itself, and everybody moves to one of the two branches, the other version of the coins will become useless and worthless. So one should be wary of buying "minority coins" or accepting them in payment.
But the tide may turn, and instead of everybody moving to XT, some miners may decide to go back to Core. If the XT slice drops to less than 50%, AFAIK the XT version will NOT reduce the block size limit again. Then there will be just one persistent chain of 1 MB blocks, but with occasional side branches that are orphaned after a few blocks. All miners will mine on the persistent branch, but when a XT miner solves and posts an oversize block, a new side branch will start, and it will be mined by XT miners until the Core branch overtakes it. Full Core nodes will only see the main branch, but full XT nodes as well all SPV clients will see the chain "stutter" as transactions get confirmed in the side branch, then unconfirmed, then confirmed again. The closer the hashpower is to a 50:50 split, the longer and more frequent the side branches will be.
In short, the states with mixed XT/Core mining are messy, and not much fun for consumers or traders. I don't think that there are significant opportunities in them.
Poll
