Centralization is the ultimate destiny of any kind of organization?

[johnyj]

Is this because human has only one brain? 

[Jutarul]

No

http://en.wikipedia.org/wiki/Swarm_intelligence

[TECSHARE]

Read "The Starfish and The Spider". IMO all centralized organizations are inherently weak in some vital ways.

[johnyj]

Not necessary to be complicated, A want to go east, B want to go west, when they act as a whole, their action will generate disagreement, and how to handle that disagreement?

Either they make agreement (centralization) or they apart (distributed), but the later means the organization actually dimissed

[myrkul]

Not necessary to be complicated, A want to go east, B want to go west, when they act as a whole, their action will generate disagreement, and how to handle that disagreement?

Either they make agreement (centralization) or they apart (distributed), but the later means the organization actually dimissed

Not necessarily. Just that it's not geographically concentrated anymore. As long as they can communicate, they can still have their organization.

[Jutarul]

the most important aspect of a decentralized system is the incentive structure. Then all you need is that everyone acts in their own best interest. The emergent phenomena is global intelligent behavior or a consensus.

Of course you're partly right - there can and will be disagreement, especially when the bitcoin protocol or incentive structure are to be altered. And that may end up in a "split". For bitcoin this is a scenario which IMHO has not been discussed very well.
However, in it's most basic form a disagreement is similar to a 51% attack. And that scenario has been discussed very well: The "stronger" party will survive, with the miners being able to choose.

The worst thing which can happen is "too many" changes "too fast". Then you may have a split into small fractions. Then a malevolent attacker has it much easier, because he only needs to attack a fraction of the total network...

[myrkul]

The worst thing which can happen is "too many" changes "too fast". Then you may have a split into small fractions. Then a malevolent attacker has it much easier, because he only needs to attack a fraction of the total network...

That's been known since Sun Tzu, if not before.

[Jutarul]

The worst thing which can happen is "too many" changes "too fast". Then you may have a split into small fractions. Then a malevolent attacker has it much easier, because he only needs to attack a fraction of the total network...

That's been known since Sun Tzu, if not before.

True. And it's actually quite a decent attack vector: Just flood the market with 100 slightly different versions of the bitcoin client which are incompatible with each other

[johnyj]

the most important aspect of a decentralized system is the incentive structure. Then all you need is that everyone acts in their own best interest. The emergent phenomena is global intelligent behavior or a consensus.

Of course you're partly right - there can and will be disagreement, especially when the bitcoin protocol or incentive structure are to be altered. And that may end up in a "split". For bitcoin this is a scenario which IMHO has not been discussed very well.
However, in it's most basic form a disagreement is similar to a 51% attack. And that scenario has been discussed very well: The "stronger" party will survive, with the miners being able to choose.

The worst thing which can happen is "too many" changes "too fast". Then you may have a split into small fractions. Then a malevolent attacker has it much easier, because he only needs to attack a fraction of the total network...

That is what I mean, even those tools that claim to be distributed like GIT, technology wise or protocol wise it is still heavily centralized, if user somehow disagree on the protocol or the way of working behind the tool, then it will split

With less and less common interest, how to hold the organization intact, that requires more thinking and work

[Jutarul]

the most important aspect of a decentralized system is the incentive structure. Then all you need is that everyone acts in their own best interest. The emergent phenomena is global intelligent behavior or a consensus.

Of course you're partly right - there can and will be disagreement, especially when the bitcoin protocol or incentive structure are to be altered. And that may end up in a "split". For bitcoin this is a scenario which IMHO has not been discussed very well.
However, in it's most basic form a disagreement is similar to a 51% attack. And that scenario has been discussed very well: The "stronger" party will survive, with the miners being able to choose.

The worst thing which can happen is "too many" changes "too fast". Then you may have a split into small fractions. Then a malevolent attacker has it much easier, because he only needs to attack a fraction of the total network...

That is what I mean, even those tools that claim to be distributed like GIT, technology wise or protocol wise it is still heavily centralized, if user somehow disagree on the protocol or the way of working behind the tool, then it will split

With less and less common interest, how to hold the organization intact, that requires more thinking and work

Don't confuse sparsity with centralization though. The fact that there is currently only one client is that only a small group of people have the will and knowledge to maintain the source code. Ideally you'd have different distributions which are all compatible. Thus no single point of failure ( like the linux OS).

[myrkul]

the most important aspect of a decentralized system is the incentive structure. Then all you need is that everyone acts in their own best interest. The emergent phenomena is global intelligent behavior or a consensus.

Of course you're partly right - there can and will be disagreement, especially when the bitcoin protocol or incentive structure are to be altered. And that may end up in a "split". For bitcoin this is a scenario which IMHO has not been discussed very well.
However, in it's most basic form a disagreement is similar to a 51% attack. And that scenario has been discussed very well: The "stronger" party will survive, with the miners being able to choose.

The worst thing which can happen is "too many" changes "too fast". Then you may have a split into small fractions. Then a malevolent attacker has it much easier, because he only needs to attack a fraction of the total network...

That is what I mean, even those tools that claim to be distributed like GIT, technology wise or protocol wise it is still heavily centralized, if user somehow disagree on the protocol or the way of working behind the tool, then it will split

With less and less common interest, how to hold the organization intact, that requires more thinking and work

Don't confuse sparsity with centralization though. The fact that there is currently only one client is that only a small group of people have the will and knowledge to maintain the source code. Ideally you'd have different distributions which are all compatible. Thus no single point of failure ( like the Linux OS).

Beat me to it. I was just about to mention Linux. They're not all 100% compatible, but the majority of those instances are the result of some company wanting to keep things proprietary.

[johnyj]

Don't confuse sparsity with centralization though. The fact that there is currently only one client is that only a small group of people have the will and knowledge to maintain the source code. Ideally you'd have different distributions which are all compatible. Thus no single point of failure ( like the linux OS).

Only a few have enough will and knowledge to work on the protocol itself, this makes it highly centralized, if majority of people have the will and knowledge to work on the protocol, there will be more and more different variants

All linux OS is based on Unix architecture, and all computers on this planet are based on binary architecture, this is also a centralized common knowledge base

[EhVedadoOAnonimato]

Either they make agreement (centralization) or they apart (distributed)

Your definitions are wrong. Agreement and centralization are not related like that. If anything, it would be the other way around:  there's agreement in an artificial decentralized order, since all participants agree to the rules and join, and there's not necessarily an agreement in centralized order, since the order may just be defined centrally and imposed to all participants. (There may be centralized organizations born out of agreements too, though).

Decentralized orders may be born of common agreements among its participants. They may also rise spontaneously (naturally), when some conditions are met, like natural evolution or the free market.

[Jutarul]

Either they make agreement (centralization) or they apart (distributed)

Your definitions are wrong. Agreement and centralization are not related like that. If anything, it would be the other way around:  there's agreement in an artificial decentralized order, since all participants agree to the rules and join, and there's not necessarily an agreement in centralized order, since the order may just be defined centrally and imposed to all participants. (There may be centralized organizations born out of agreements too, though).

Decentralized orders may be born of common agreements among its participants. They may also rise spontaneously (naturally), when some conditions are met, like natural evolution or the free market.

second that. Agreement comes from the need to cooperate. Another example is language. Why do you think people settle on only a few languages globally? Nobody "tells" them. It's doing you no good to learn languages which nobody else speaks....