Quote from: Come-from-Beyond on January 13, 2015, 09:53:41 AM

Today I saw http://bytemaster.bitshares.org/article/2015/01/13/Decentralization-of-Nxt-vs-BitShares/ and became interested by the title enough to spend some time on reading. The article is related to http://bytemaster.bitshares.org/article/2015/01/07/The-Most-Decentralized-Proof-of-Stake-System/ and the both analyze the same phenomenon (decentralization), so I will treat them as a single article.

I'd like to comment some things mentioned in the articles.

...

PS: It would be great to see comments on comments...

It's great to see you here, and I appreciate the civil tone of your comments!
(Its like a breath of fresh air.)

I have notified Bytemaster that you have made some serious comments on his article and I am sure that he will study them closely, and if necessary correct the referenced article. We certainly don't want to misrepresent anything about the product of anyone else's blood, sweat, and tears!

I assure you that no attack on NXT is intended beyond an attempt to respond to others here who requested a concrete example. We are in no way suggesting that NXT is not a viable product with a bright future.

There's a lot of discussion in these threads and the wavefront keeps bouncing around between three of them:
https://bitcointalk.org/index.php?topic=920621.0;all (now apparently locked)
https://bitcointalk.org/index.php?topic=916696.40
https://bitcointalk.org/index.php?topic=913075.0;all (this thread)

So I don't expect you to have come across the context or motives behind the subject article.

I think we have adequately dealt with the more reckless postings in these threads, but there is an interesting recurring debate (when you remove all the clutter) about the meaning of the word "decentralized". Apparently one opinion is that the connotations of the word in this industry have moved it away from its dictionary denotations and that further evolution of its meaning in response to ongoing technological innovations should be resisted.

Our point has been that there are multiple valid ways to move control away from the center and that having all stakeholders select from a high-quality set of candidate block signers based on their established reputations has some design advantages over randomly selecting signers from a much larger group of candidates with unknown reputations. (At least for some systems with certain additional design objectives and target applications.) Signing authority is uniformly distributed among the 101 most respected members of the community without regard to the size of their stakes. This can change every 10 seconds and everybody has a chance to participate based on merit, not chance. Signers have strictly limited power. They can either faithfully do their job or be detected and immediately fired.

The key is recognizing that there are multiple design degrees of freedom we can play with. We need enough scalable signing nodes to be fault tolerant and a decentralized way to select them. Both NXT and BitShares accomplish this objective in different, valid ways. Whether each stakeholder does the routine and transparent signing work on their own computer or on a computer from someone with a vetted reputation they are still selecting which computer they want to have do it. I know I'd certainly rather pick somebody I trust to do the signing than take responsibility for that technical specialty myself!

All "trustless" systems allocate residual trust somewhere, usually by default. We have made it explicit. Making it explicit in turn gives us a unique by-product asset - trustworthy nodes selected by all stakeholders. Our designs then leverage these assets to assign them other functions that benefit from established trustworthy reputations, where any breach of trust can be detected and instantly removed by all stakeholders as well.

A recent article I wrote with Bytemaster attempts to clarify this distinction by clearly highlighting the roles of decentralization, scalability, and fault tolerance in designing robust, profitible, and incorruptible systems. We would value you feedback on it.

Decentralization, Scalability, and Fault Tolerance of Bitcoin