Quote from: smooth on January 07, 2015, 08:54:41 AM

A gazillion hashes/second doesn't make the network secure and in fact the number of hashes is completely irrelevant. What matters for the most part is the amount of electricity used and secondarily the cost of the mining equipment. For example 100 watts buys you something on the order of 200 GH/sec of SHA256D. The same 100 watts would buy you perhaps 300 hashes/sec of (to pick one with which I happen to be familiar) CryptoNight. These are approximately equal in value for mining; 200 GH/sec of one algorithm is not 2/3 of a billion times "more secure" than 300 H/s of another. These units are incomparable.

The current bitcoin network is about 150 megawatts. That's something on the order of a 1-2 million PCs using a CPU algorithm. At least until specialized hardware and mining farms developed, a bitcoin forked to a CPU/GPU algorithm would look a lot like a million bitcoin users all mining on their computers. Sounds pretty darn secure to me.

(Is this your "technical objection"?)

You are assuming that 2 million bitcoin users would all rather upgrade to the "rebel" chain and start mining on their own CPUs rather than accept the cartel's proposed change to the protocol. (Can you see the "belief about how the community would react" there?) It is like saying that if the banks decided to raise their fees, all their clients would take their cash out and create a banking system of their own, out of their garages and kitchens.

But let's say that all users were to feel like that. First, the number "2 million" depends on who you count as "bitcoin user". (I gather that the number of full nodes is less than 10 thousand, isn't that so?)

As of last september, there were only 650'000 bitcoin addresses in the blockchain with at least 0.1 BTC in them; and less than 160'000 with at least 1 BTC in them. Addresses are not people, of course, and many people who own bitcoins keep them all in ~~MtGOX~~ ~~Bitstamp~~ Coinbase or some such place; but wouldn't you agree that someone who has more than a uh-tried-it-once interest in bitcoin should have at least one blockchain address of his own with at least 1 BTC in it?

So, if we take that criterion as a cutoff, we are talking about 200'000 real users at most. Now, how many of them would have the knowledge, time, patience, and computing resources to downlad the "Red Button" mining code and start mining? How many will put up with the extra load on their PCs and laptops to secure their 2.73 BTC? 10% maybe? ("belief about how the community would react" again)

So, even if most of the users hate the cartel and choose to upgrade their wallet software to the Red Button protocol, there may be perhaps only 20'000 PCs and laptops mining the rebel chain. At (guess) 300 W per machine devoted to mining, 24/7, that would be 6 MW securing the Red Code fork. If the cartel wanted to jam it too, how much cloud computing power would it need to rent to match their Red Button hashing power, and how much would it cost?

But let's say that a substantial number of users do join the rebel ranks, and the cartel leaves the Red Code chain alone, or fails to jam it. So then there would be two two competing and incompatible coins, SunniCoin and ShiaCoin, both claiming to be the real Bitcoin.

One coin uses practically the same protocol, with a small change to the halving schedule, and claims to have a large state-of-the-art mining network comprising all former bitcoin miners (mostly turned off at present, because of the lower coin value would probably negate the doubled reward).

The other runs a protocol that is mostly similar to Satoshi's, but is incompatible with all mining equipment, and has a much smaller network (whether measured in MW, or in CPU computing power).

Everyone who owned N bitcoins right before the split, whether he likes it or not, would own N SunniCoins and N ShiaCoins just after the split. However, the two blockchains would inevitably diverge, therefore transactions in one cannot (and should not) be automatically mirrored in the other. Indeed, those users who are not absolute purists will hoard, sell, or spend both coins independently, for whatever price each would fetch. To do that, they would have to duplicate their bitcoin wallet at the time of the split, and install both versions of their wallet software. Most users will probably do that eventually (the alternative is pure loss), unless one of the coins dies out first.

Note that both coins will start out with the same set of owners, so neither can boast of having most of the users. A user who wants SunniCoin to die can only refuse to be paid in SunniCoin, and spend, sell, or burn all his SunniCoins. So, paradoxically, if most users want ShiaCoin to flourish, there will be a surge of traffic in the SunniCoin blockchain, and vice-versa.

How would prospective crypto investor choose between the two? If someone was looking for a venture to invest his retirement money, which would he choose: a restaurant in Donetsk, or an apparel store in Northern Iraq?

Obviously both coins will see their price collapse, and their combined marketcap will be much less than the bitcoin marketcap before the split. It is as if bitcoiners woke up one day to find that each bitcoin they owned had been swapped for one Litecoin plus one Dogecoin.

So, the RedCoin solution would be a near suicide by the users, even if they had support for their coin in all payment processors and other services. It would break the trust in both coins, not only among ideologues, but also among the majority of the "materialistic" users. Large holders, fund investors, the service companies, and the VC investors who put their money into them would together lose a billion dollars or more. Ditto for any other solution that splits the blockchain into two independent chains.

Now rewind the movie to before the split. You own an exchange or a payment processor. You must decide whether to come out in favor of the cartel, or of the announced rebellion, or just avoid taking sides. In any case, you must handle the two future forks as two distinct coins. If you do not want to take sides, you should provide support in your service for both coins, each with its own set of accounts (and its own ledger, order book, or whatever). Note that each bitcoin that your clients had deposited will become one SunniCoin and one ShiaCoin, both under your control.

The alternative is to provide support for only one of the coins, in an attempt to push your clients towards it. But if you provide support only for ShiaCoins, you must still let your clients withdraw their SunniCoins, otherwise you would be stealing them.

Now, if most of the bitcoin users submit to the cartel, your losses will be minimal, assuming that you provide support in your service for the cartel's coin. In any other case, you will lose millions of dollars, even if you provide support for the rebel's coin. What would you do?