What is wrong with the goal of decentralizing development across multiple competing implementations?
Nothing, but that is absolutely not what is happening. I know this is your new talking point (a very tired strawman btw) Peter but it doesn't reflect on the situation at hand:
A totalitarian power grab for the governance of Bitcoin. Not the proposition of a new implementation but an attempt a hijacking the consensus code behind political motives
Please don't turn into Stolfi you are a valuable asset to this community when your head is in the right place
#1. I disagree. What I see is demand building for bigger blocks, Gavin and Mike providing a solution, and people moving towards that solution. This is the free market at work. If we had several more competing implementations this "consensus process" would be more efficient.
#2. The only talking point the small block side has left is "big blocks = centralization" (which I disagree with but can't prove it). What I find ironic is that they simultaneously fight hard to keep centralization in the most heavily centralized aspect of Bitcoin: development!
If we need centralized development to keep Bitcoin decentralized then we've already lost.
Let's increase the block size
and decrease centralization by showing our support for a wider variety of Bitcoin implementations!
1. The very fact that practically no one is actually moving towards that solution should tell you all you need to know about the morality and quality of it. A very bad solution is not much better than no solution at all. Especially as it creates dissent and enormous loss of productive time within the developer community.
2. You absolutely don't understand or are being willingly misleading about this "centralization" issue. There is quite simply no other choice but for us to support a centralized (read unique) consensus code. That's pretty much the only way Bitcoin works. It happens that the core developers have historically been the one trusted with maintaining this code and updating it. Several
implementations have been built around this consensus code. Most of them have little support for very valid reason: their implementation is generally considered less tested and therefore potentially less secure than core implementation. Now should we blame core for attracting the most competent developers in the space? Would it be rational to ask of them to each start dividing their work between different implementations just for the sake of "decentralization"?
The centralization issue you refer to is nothing more than a lack of man power. That is, only a scarce amount of people are reliable and technically able enough to handle the highly fragile development of Bitcoin. It is no wonder the guys currently leading core are some of the world's most advanced experts in their own field. This expertise cannot be easily replaced or dismissed "because decentralization". It is absolutely unproductive and irresponsible to try to advance decentralization of Bitcoin development by encouraging incapable people to start messing around with their own implementations and risk breaking consensus.