There has been an ongoing debate over the relative decentralization of various projects. I have made several posts where I analyse things from
both an economic and mathematical perspective and draw various conclusions about the theoretical decentralization of the various protocols. Today
I discovered that Nxt has already published all of the information necessary to prove my predictions correct.
There is a vast difference between what something is and what it is used for. Money or currency is an attribute of how something can be used, but does not give us a clear understanding of what something is. Anything can be used as money provided it is sufficiently durable, divisible, fungible, transportable, and resistant to counterfeit. Cigarettes are a commodity but can be used as money. Gold and Silver are metals (also a commodity) that have been used as money. So what is Bitcoin when it is not used as money and why does it have value?
Bitcoin is all about decentralizing the control over money through objective consensus on ownership. What often gets lost in the discussion is the
difference between decentralization, scalability, and fault tolerance. Bitcoin is an example of a decentralized system whose scalability
is limited by the power of an individual node rather than the combined power of all nodes. This means that a system’s ability to become
distributed for fault tolerance, decentralized for control, and scalable for performance are not linked together by the number of nodes. In
fact we could say that attempts to combine these roles via the same mechanism will carry with it the combined limitations rather than the
combined benefits. Hence economies of scale and free market competition will tend to minimize unnecessary fault tolerance (redundancy) while centralizing for
performance. If we are not careful this will result in centralization of control.
When it comes to crypto currencies there are no take-backs. A single mistake could send thousands of dollars to the wrong person never to be seen again. Crypto currencies go to great lengths to minimize human error. This normally takes the form of a checksum tacked on to the end of the payment address. With Bitcoin this checksum doesn’t impact the user experience in any significant way because all addresses are essentially a random string of characters that no sane human would ever attempt to type by hand. In BitShares we have attempted to make the whole user experience easier and replace random addresses with human friendly names, but in the process introduced a new point of failure: typos. To reduce the risk of typos we have implemented a new feature that we call your SafeBot.
There have been many books that I have read over the years that I have found incredibly useful in helping me understand the world around me. Each book has helped me on my journey to who I am today. I will be adding to this list as I discover new books that move me.
The crypto-currency industry has been abuzz ever since Wired published an article entitled “Overstock’s Radical Plan to Reinvent the Stock Market With Bitcoin.” Patrick Byrne, CEO of Overstock.com, has seen firsthand how Wall Street has been corrupted to the core, and he’s eager to do something to fix it. He has identified two primary sources of corruption: centralized clearing (exchanges) and fractional reserve banking. These systems enable unscrupulous insiders to sell shares in companies that don’t exist while allowing investors to exploit loopholes in stock settlement, such as naked short sales, harming public companies and the economy as a whole.
The crypto currency community is full of people making wild claims about the relative centralization of various projects. Calling something centralized is like calling someone racist; it starts to lose all meaning unless you are extremely careful in how you define it. Decentralization has become a kind of religion, when in reality it is merely a tactic to deal with certain kinds of corruption and attacks. Today I would like to document how I measure decentralization of various crypto projects.
There are many arguments against Proof of Stake, but none of them is more persistent than the fear of the so-called “Nothing-at-Stake” attack. Under this attack it is claimed that block producers can produce an arbitrarily long alternative chain at any time and then use this attack to “kill” the network. All that is required is for people to buy up old private keys that at one point controlled a large amount of stake and now no longer do. They have nothing at stake and thus will sell the keys cheaply. This so-called attack is a real stretch compared to much more realistic attacks against Proof of Work. It is high time this argument be let out to pasture.
A couple of weeks ago we launched the initial version of DevShares without consulting with the community first. At the time I thought the allocation shouldn’t be a big deal because it would be a low value, highly unstable, coin with no intention of becoming a competitor. It became apparent that I was mistaken about the symbolic importance of the allocation. Today I would like to apologize for any confusion and the mixed signals. Then I will reveal details of our latest effort to launch our DevShares network. If you are unfamiliar with what DevShares is, checkout my post on The Value of DevShares.