These modifications don't seem like difficult to do in 4 months, right ?
Making the modifications is not enough. You need to upgrade every bitcoin node in the world as well.
Not really. If you keep your old node running, you copy the the old fork, if sufficient miners go with the old fork to still make some blocks. If you download the new node, you copy the new fork.
And lose bitcoin. How well do you think that will be welcomed by the greater community, and how much would you trust a currency whish did that?
You never lose bitcoins because you run a non-mining node. The only things that matter is what is recorded on the block chain(s). Whether your local copy gets fucked up or not doesn't matter. The only thing you can have as an accident, is that your old software makes a funny transaction that is nevertheless accepted in some way by the miners and put in a chain, but is in a way screwed up that you cannot use its outputs any more.
But by running an old node, you never "lose bitcoins". And if bitcoin forks in two chains, you have your former coins on both of them. Maybe you don't SEE them, because you're not using an appropriate node, or an appropriate wallet, but they are there. If my secret keys are safe, even if I wipe my hard disk and erase my full node, I didn't lose my coins. (and the keys are in my *wallet*, not my *full node*: hell, at home I have an empty full node, and my bitcoins are in a light wallet on another computer which connects to my full node, so whatever I do to my full node doesn't affect my coins).
If it was simple to change the hard economic consensus parameters, like block size, inflation rate, time between blocks, POW algorithm etc, it would have happened several times already. It doesn't, because people want bitcoin to be a secure store of value.
It doesn't, simply because of the mechanism of immutability, which, however, can break down if centralization occurs and there's a collusion of more than 50% of the consensus (= hash) power over a change.
But at least I'm happy that you consider block size just as well a hard economic parameter as inflation rate. I think that the block size limit as an economic parameter, introducing scarcity of transaction room, was a stupid thing to do in bitcoin's design, but so is its inflation rate. So, bitcoin being designed as a system with a scarce and finite number of coins, I don't see the problem with bitcoin as a system with a finite and scarce number of transactions per unit of time. I have to say I think the economic model of both is stupid if the idea was to make a currency, but then, that's how bitcoin was designed, and I think that is the way it should live its life. The economic design looks more like the one for "exclusive famous paintings" which are rare to come by, and difficult to transact, in other words, a kind of highly speculative and not very liquid asset with high price that is rarely moved, and only to move big amounts of value (not a currency at all, but a "settlement layer for rich guys doing things where fiat cannot go").
But in all of this, you don't even need to run a node. You can just connect your light wallet to one of the miner pool nodes.
Yeah, or just use PayPal if you want to trust a thrid party. Actually I think PayPal is more trustworthy than the miners. That's why I chose to run my own nodes.
The point is, you can ask for the books of PayPal, or you can ask for the books of the miners. That's what you do when you use a full node. But you cannot change them, and there's only one book out there. If you think that PayPal has been cheating in the books, you could go to a judge. If you see that the miners have been cheating, I don't think you can go to a judge. You can just curse them, and that's it. If the one book that is out there is not to your likings, what are you going to do about it, apart from shouting, cursing, trying to tell everyone not to use bitcoin because it is a scam ....
Suppose you hold, I don't know, 10 000 BTC, and you see that the miners are cheating. Suppose that you see suddenly that they are giving themselves a 200 BTC block reward. Your node stops. There's no other chain around. Are you really going to shout that bitcoin is a big scam and that the miners are screwing everyone ? If you run an exchange, are you going to stop all transactions of your customers ? Or are you silently going to modify your node software to be able to transact *your own 10 000 coins* and, maybe, sell them before the shit hits the fan ?
This won't be a problem, since old nodes don't generate segwit addresses. You can pay him with your segwit coins, and it is secure.
Ah, I didn't know you could go back from a segwit address back to a legacy address. How can the old node check that transaction, given that he doesn't have the witness data ?
Suppose that I had coins on a legacy address A1. I transfer them to my new segwit address S1. Now, Joe, running an old node, has address A2. Can I transact coins from S1 to A2 ?
But, suppose now that I had coins in S1, and I pay Jack, running a new node, in S2.
I could try to spend S1 to A2, because Joe, with his old node cannot see my transaction from S1 to S2.
But of course, the *miners* will not accept my transaction from S1 to A2, because that would be a double spend. In other words, Joe, with his old node, cannot see that I'm doing a double spend, and would cheerfully accept a chain with a spending from S1 to A1 (if this is even possible ?), but he TRUSTS THE MINERS that they won't allow that.
What's the point for him to run his old full node, and not a light wallet connected to a miner node ?
You may argue that segwit is a cleaner way of doing things, but there is no need to hard fork for it. In fact it will be very stupid to hard fork for a simple change like that. P2SH was a much more intrusive change, and it was done by a simple flag day activation.
The point is that if you do a radical change in the protocol, you fork anyhow. There' s no good reason to keep backward compatibility with software that doesn't understand the new protocol but simply "allows it". The coin after is not the same coin as the coin before. The protocol is different. The only thing that is the same, is the ownership of coins.
And a clean hard fork would also allow people to "not be tied to backward compatibility". Many crypto currencies have such a policy. There's a lot of clumsiness in the requirement of a soft fork that disappears with a hard fork. For a radical modification like this one, a hard fork is much cleaner.
Use one of the scamcoins, if you want an insecure coin which hardforks all the time. Don't think you will be able to convince all bitcoin users that would be a good idea, and then you have two coins, disruption and big losses for everyone.
This is what I call religion. If you talk about "insecure" and "scam coin", that's not rational. Hell, I'm even sure that you can change the inflation rate in bitcoin with a soft fork too. If you would allow every first transaction from legacy to segwit address to be spent not once, but twice, you'd double in one go, all bitcoin wallets that switch to segwit. With a segwit soft fork because it is a new protocol that is "invisible" to the old one, so a soft fork.
The whole "leading argument" in this whole business is the irrational belief that non-mining full nodes have any decentralization value, and that old nodes with old node software are important. Both of these notions are entirely wrong, but they are the fundamental argument on which all of this dispute is based.
Don't try to tell me this is wrong. I run an exchange. A small one, about 1 million USD in monthly volume now (times two, if you count buy and sell separately). Quite often when people sell coins to me, it first takes them ages to sync their old Bitcoin QT node. There are thousands of old nodes out there, and people who run them.
You could also just run a full node for your customers, and tell them to connect a light wallet to your node. Or to any other node. Ultimately, to the source of all that data, which is a mining pool node. Because all that node does, is copy what the miners produced. If there's only one block chain out there, and the block headers link together and the last one is "up to date", then cryptographically you know that you have the headers of the sole chain out there.
A light wallet does all the checking that is needed for the transaction at hand, on the single chain that is in any case out there.
I would consider it "safer" to use an up-to-date light wallet, than a legacy full node. And in any case, there's only one "book" out there, whether you like it or not.
A (light) wallet is like a web browser, and a full node is like a proxy server. The original data server is the backbone of miner pool nodes, that make the unique block chain, to take or to leave.
