I do apologise. This again turned into a far longer post than I imagined, but I'm trying to consider all angles. All comments are highly welcomed:
And we crash again, just as people start complaining about transaction delays, like clockwork.
Why do you think it is a terrible idea? We currently have an irreconcilable ideological difference between those who think that bitcoin should be a settlement system and those who think that we should scale while we have the subsidy so that security is provided by the number, not fee amount, of transactions.
I do not think these differences can be reconciled with gmax having publicly stated that he will continue working on/maintaining the 1mb chain even if everyone else moved over and he probably would have some perhaps 10-20% support. As such, if bitcoin scales through a hard-fork, there would be two chains in any scenario, with one of them eventually having a pretty low value, but perhaps more than litecoin as it probably would take the second place.
I find the current situation to be highly undesirable as we are effectively begging gmax to change his mind while he has unambiguously stated that not only will he not change his mind, but he will resist until the bitter end. Nonetheless we are still waiting for him to change his mind, pressuring him, and as any emperor, he entertains us at times by giving us the minimum increase of 2x and even that in some 6 months.
The miners on the other hand are like deer caught in the headlights and therefore have defaulted to no action claiming they are not the jury. Which means that we are basically going in the way of a settlement system with king gmax making proclamations, responding to public pressure by a minimal increase, and doing his emperor things.
With theymos having effectively taken control of the communication channels and seemingly impossible to remove regardless of how clear his scamming/thefts are made to the admins, with some miners now going as low as begging for 2-4-8 and even that denied, thus defaulting to no action, with gmax dead set on a settlement system, we can cry all we want about censorship, we can beg all we want for an increase, but we are effectively asking for permission from centralized entities who will do what they want as they always do and that is far from bitcoin.
Obviously, forking with a changed algo is not ideal, but I can see benefits. It would decentralize mining in an instant, allowing everyone to mine with their cpu, which would be a huge sell within and outside the bitcoin community as this would be the first time that the network effect is maintained while cpu mining is "liberated" to all. Security of course would be lower to begin with, but only for a few months as the hashrate race does its job. There would be a free, not fee, market choice and, most importantly, bitcoin would scale.
One would think that the only necessary step for the miners to take a positive step is a credible pre-announcement of ST (or whatever name such as OG, Classic, 3.0, ST for satoshi level transactions etc) as at that point things can become existentially unpredictable for the miners. For the hodlers there would be as good as no change as the value (combined or otherwise) would remain the same or increase in the medium and long term. For speculators there would be a bonanza if it gets to post pre-announcement stage. For bitcoin there would be incredible strength in showing to all, from Chinese peasants to USA presidents, that we do not ask for permission in this space.
I can see two ways of doing it. The first and ideal, far superior, way would be for someone credible to pre-announce ST and then exchanges to publicly announce that they will move over to it and that if they do move they will not go back so as to prevent QT miners from killing the chain by implementing BIP101 after the exchanges have moved thus creating utter chaos which is easily avoidable by exchanges proclaiming that they will not go back if they move. This would place miners on high alert and create the necessary momentum and pressure for a positive, rather than passive, decision. One would think that within days of such credible threat, a miner consensus would be achieved with a reasonable proposal acceptable to the supermajority such as perhaps bip101, but starting at 2mb together with SW but as a hardfork.
If instead the miners would still not reach a decision after the economic majority has made its will clear, and therefore has placed their entire livelihood in a credible and real danger, then that would show that there are some serious and fundamental problems in the bitcoin ecosystem and therefore that the benefits of the changed algo fork would far outweigh the dangers. I however consider this to be incredibly unlikely as miners are fundamentally profit motivated entities who have invested millions in their hardware and therefore would follow the users and if not the users are entitled to fork them off.
The second way would be more convoluted and perhaps unlikely to work, but on the other hand does not require anyone’s permission and allows for a wait and see and real life testing approach in regards to the implications of far lower day1 security. Someone could pre-announce the ST fork, but this time exchanges do not state that they will move over. In this case ST would in effect be an altcoin, but with the huge benefit of having bitcoin’s network effect as pre-fork coins can be used in either QT or ST chain.
The problem is that until exchanges announce they will move over, it is unlikely it would gain any credibility, thus having far fewer users and thus much lower value, but still higher than litecoin’s due to maintaining bitcoin's network effects. The further problem is that as 100k transactions are made a day, for each day the fork continues more and more users are unable to interchangeably move their coins, thus making it less and less likely that exchanges would move as many more users would be against it due to their coins not being interchangeable. They could however always just sell their QT coins and buy ST ones.
The further problem is that until exchanges move it can be instantly killed by QT miners adopting BIP101. Thus making it a gamble, but probably a profitable one, at least in the short term.
The second option does have benefits however. It would allow the fork to be live tested with only gamblers/speculators on it, it would allow for a smoother transition, it would allow for free market voting and it would exert huge pressure on QT scaling and give exchanges the option of moving to ST, thus conclude the debate. Furthermore it is permissionless, so if there are profits to be made then you'd think someone would go ahead with it for good or ill.
Both options have drawbacks, with the first option’s main concern being day1 security. However, I cannot possibly imagine a lack of quick conclusion to this debate if exchanges stated that they will move to the ST fork and at this stage of the debate I do not think miners can credibly call them bluff. If miners do not take a positive action even then, then perhaps there is no choice, but to fork them off and without this credible threat I can not see miners taking a positive action after their comments at the workshop, therefore perhaps the economic majority has no choice but to offer this path.
On a side note, I find it interesting that no one has forked the chain with the intention of creating and maintaining a concurrent chain. Doing so is in effect creating a sidechain and seems to me superior to creating an altcoin through a genesis block, so I find it curious that no one has ever tried considering the huge potential profit and permissionless nature, which makes me wonder whether I am missing something.