Bitcoin Forum

The purpose of this thread is to discuss the various ways in which Bitcoin is centralized, and the risks imposed by each form of centralization.  To start the discussion, here is a graph that illustrates the centralization in nodes, mining and development:

https://i.imgur.com/bfBeGZb.gif

Source: https://www.reddit.com/r/bitcoinxt/comments/3i37m6/centralization_in_bitcoin_nodes_mining_and/

The dev team is necessarily centralised. When centralised is the only ideology that works, I'm in favour.

Pick your centralising influence: amorphous Core team or Mike Hearn's self described "benevolent dictatorship". Benevolent in respect of who, exactly?

Can you explain why you believe a model with several competing implementations of the protocol wouldn't work?  In fact, with the emergence of BitcoinXT it looks like it is beginning to work.  The consensus critical code would of course need to be compatible between implementations; however, if a change was desired by the community (e.g., increasing the block size limit), each implementation could attempt to solve it "their own way" and then the community would decide the winner by switching to the implementation they favoured.  Then, to retain some portion of their previous user base, the losing implementations would adopt the same change to the consensus code to prevent their clients from forking off from the longest proof-of-work chain when the change goes into effect (at some future time similar to BIP101).

Is this a way to give the community more power in exercising their free choice?

Well, yes. A core dev team leads development...

...until they betray the wishes of the
community, at which point other developers step in. 

That appears to be the current reality.
 
I do think several implementations are healthy though,
but you're probably right that changes need leadership.

Well, Mike and Gavin just proved this to be wrong.


It actually doesn't. Otherwise XT would not have come into existance in the first place... XT is the results of the Core team ignoring a large part of the market that are asking for scalability faster than what they were projecting which IS a problem due to centralization.

Linux operates under that model and now is, probably, the most used operating system in the world, if you count Android as being Linux...

Linux doens't need a consensus to prevent an potential colapse though...

Yap, the implications are quite different, there are a shitload of Linux forks and that's ok, the more the merrier.

Bitcoin developing/governance model is a concrete problem, it's curious how it only became obvious by now.

Actually IMO the problem that is not being realised is that we don't need "one blockchain".

Once people can get over that idea then true decentralisation can start to take place (and yes that is a hint at what the CIYAM project is about).

Of course it is. Competing implementations is the only way to make it a market choice. In some years we will wonder why this was ever not obvious.

You've been hinting this alot, can you stop it already? until you got something to show..

I agree.  In fact, I see this as the Bitcoin Consensus mechanism playing out exactly like it should.  The Bitcoin white paper says that nodes express their acceptance of a block by building on top of it; the longest chain becomes the correct chain.  Giving people the choice between competing implementations just makes it easier for them to express their acceptance of, for example, blocks larger than 1 MB. 

In hindsight, it will be seen as ironic that so many fought for centralized development in order to keep Bitcoin decentralized.