Seeking Consensus on Consensus - DPOS or Delegated Proof of Stake and the Two Generals' Problem

在共识中寻求共识-DPOS(委任权益证明)和两军悖论

作者：Ian Grigg

翻译：Gavin

校对：Lochaiching

Laying down the rails for a high performance financial blockchain-based ecosystem is well understood if controversial because there are a number of approaches - centralised, decentralised, un-permissioned, walled garden.

为一个高性能的金融区块链生态系统奠定基础肯定是备受争议，这很好理解。因为这里有很多方法——中心化的、去中心化的、未经许可的、封闭式。

Having chosen all that, we still have that one nagging itch of how to deal with race conditions. No matter the design, there’s always something you want to do and something someone else wants to do where these two things can’t both happen. Or, it’s cast in the sense that you want something to happen, and you want someone else to know this is going to happen too, known in the computer science worlds as Two Generals problem:

除去以上提及的这些，我们仍然有一个关于如何应对竞态条件的烦恼。不管怎么设计，总有一些你想做的事情，还有一些其他人想做的事情，而这两件事不能同时兼得。或者，在某种意义上说你想要一些事情发生，并且你还想让别人知道这事情将要发生，在计算机科学领域中被称为“两军悖论”：

Two armies, each led by a general, are preparing to attack a fortified city. The armies are encamped near the city, each in its own valley. A third valley separates the two hills, and the only way for the two generals to communicate is by sending messengers through the valley. Unfortunately, the valley is occupied by the city's defenders and there's a chance that any given messenger sent through the valley will be captured.


“Positions of the armies. Armies A1 and A2 need to communicate but their messengers may be captured by army B.”



While the two generals have agreed that they will attack, they haven't agreed upon a time for attack. It is required that the two generals have their armies attack the city at the same time in order to succeed, else the lone attacker army will die trying. They must thus communicate with each other to decide on a time to attack and to agree to attack at that time, and each general must know that the other general knows that they have agreed to the attack plan. Because acknowledgement of message receipt can be lost as easily as the original message, a potentially infinite series of messages is required to come to consensus.



The thought experiment involves considering how they might go about coming to consensus. In its simplest form one general is known to be the leader, decides on the time of attack, and must communicate this time to the other general. The problem is to come up with algorithms that the generals can use, including sending messages and processing received messages, that can allow them to correctly conclude:



“Yes, we will both attack at the agreed-upon time.

Allowing that it is quite simple for the generals to come to an agreement on the time to attack (i.e. one successful message with a successful acknowledgement), the subtlety of the Two Generals' Problem is in the impossibility of designing algorithms for the generals to use to safely agree to the above statement. (cite Wikipedia(https://en.wikipedia.org/wiki/Two_Generals'_Problem#Definition))”

两名军队由分别由其将军领导，准备袭击一座要塞城市。 军队在城市附近扎营，每个都在自己的山谷中。 第三个山谷将两座山谷分开，两位将军交流的唯一途径就是派遣信使穿过山谷。 不幸的是，山谷被城市的防御者所占据，并且任何被送往山谷的信使都有可能会被捕获。



军队的位置：军队A1和A2需要通信，但他们的信使可能被军队B抓获。



虽然两位将军同意他们会进行攻击，但他们还没有就攻击时间达成一致。两位将军要是不能同时攻击城市，那剩下的另一支军队也会随之覆灭（即只有两支军队同时进攻才能获取胜利）。因此，他们必须与彼此交流以获得进攻时间上的共识，并且两位将军都必须知道对方明白双方已达成共识，原因在于确认收到对方消息的这一条信息也有同样的概率会丢失，这就导致：两人需要就一个潜在的无限条信息组成的信息流需要达成共识。


这个思想实验包含：考虑他们该怎样达成共识。在这个最简单的形式之中，已知一个将军是领袖，决定进攻时间，并且必须将这个时间告知另外一位将军。现在问题在于我们需要一套算法能够让将军用来发送消息和处理已被收到的消息，并需要他们正确的涵盖以下内容：



是的，我们将在已达成共识的时间发起进攻。


要做到这一点是很容易的（即一个成功的消息和一个成功的确认）。两军悖论的微妙之处在于设计一套真正安全并能满足以上条件的系统是几乎不可能的（引自：[维基百科](https://en.wikipedia.org/wiki/Two_Generals'_Problem#Definition) )

Also known as the coordination problem. In blockchains we call it the consensus problem. In the financial cryptography world, it’s the double spend problem, and in databases, atomicity. Which is to say this is a fundamental problem in all of computing science, and it’s not new just because, ya know, blockchain.

在区块链中，我们称之为共识问题，也被称为协调问题。在金融密码学领域，被称为双重支付问题，在数据库中被称为原子性。也就是说，这是所有计算科学的一个基本问题，它并不新鲜。

Let’s work through the evolution of this problem.

让我们来研究一下这个问题的演变过程。

Centralised Double Spend Protection

中心化双重支付保护

The original mechanism in financial cryptography is the simple client-server or trusted third party (like SOX), which is to say that the issuer of a double-spendable value like a coin runs a single server that mediates the double spends. Typically, the requests are queued up on a first in, first out (FIFO) basis, which is standard in databases these days.

加密金融中的原始机制是一个简单的客户-服务器端或受信任的第三方（比如SOX，译者注：一种支付系统），这就是说支持双重支付的价值发行方（译者注：这个发行方是一种货币）仅仅依靠一个服务器在双重支付中进行调解。十分典型的是，请求遵循先入先出（FIFO）排队方式，这在今天的数据库已经是标准了。

Blinded Cash

盲现金

One notable variation of centralised double spend protection was the blinded signature over coins, invented by David Chaum in the late 1980s (Chaum, Achieving Electronic Privacy). The blinded coin was independently usable (verifiable) as a monetary token, but like all data, Alice could hand it to both Bob and Carol at the same time. The solution then was for Bob to push the coin back to the issuance server, known as a mint, and ask for a freshly signed one. Carol should do the same, and be faster!

一个中心化双重支付保护的显著变种就是对货币的盲签名，这个概念由David Chaum于上世纪80年代末期发明。盲货币作为一种货币token可以单独被使用（可验证），但是就像所有的数据一样，Alice可以将其同时给Bob和Carol。对Bob而言，针对这种情况的解决方案就是将其收到的货币退还给发行的服务器，也就是印钞厂，并索要被重新签过名的另一份货币。同理，Carol也可以这么做，甚至以更快的速度。

“盲签名”注释，校对者注

Now, the special trick of the blinded signature on the token was that when pushed through a modified form of RSA, it could morph to create a new signature that the issuer had not created, but was still valid and therefore could identify a real coin to the issuer. Blinding therefore meant the issuer couldn’t track what Alice did with her coins, a very valuable thing! A further trick with eCash was that if Alice did try to spend the coin twice, the mint could use some crypto to combine both erstwhile spent coins, strip out the psuedonymous protection, and reveal who the naughty girl was.

现在，盲签名在被用于Token上时存在一个特殊把戏：当Token完成一次被修改的RSA加密算法时，其能够创造一个之前未被使用过的全新签名，但这个签名却依然有效，因此这个货币依然是被承认的。因此，盲签名的“盲”字表示，货币的发行者不能追踪Alice用这些货币干了什么，这显然是十分有价值的！另一个关于电子现金的把戏是如果Alice确实尝试两次使用同一笔货币，造币方可以使用加密手段来将这两笔被消费的货币合并，去掉匿名保护，并揪出罪魁祸首。

盲签名案例，校对者注

I include mention of blinded double spending partly because it is an example of a very complicated scheme to catch double spending, and partly because it started our field of financial cryptography - invented back in the mid 1980s and built in the mid 1990s.

我涉及了部分关于盲双倍支付的内容，因为这是一个非常复杂的用于管制双重支付的系统中的一个例子，也因为这是我们诞生于上世纪80-90年代的金融密码学领域的开端。

Voting

投票

Replicated servers became the in-thing typically for purposes of reliability. For example, the early NASA space shuttles had a voting ring of 3 primary IBM mainframes (and a couple of standbys). On every important act in a voting circle, a majority would win, and a minority could be disconnected and replaced. Early simple majority voting schemes proved to be a lot of trouble, and now the ruling buzzwords are Paxos and PBFT (practical byzantine fault tolerant), but do note that behind them there are lashings of Lamport, theory, bickering Byzantine Generals, PhDs, papers and Turing Prizes, oh my!

多服务器的选择变得流行主要是为了追求可靠性。比如，早期NASA的航天飞机就有以三个IBM主框架形成的投票环（以及一些备用框架）。在每一个投票系统中，都是占大多数的获胜，其余的小部分会被断开连接，并被取代。早期的建议投票系统事实上被证明有许多不足，而现在在行业中流行的关键词是Paxos（一种基于消息传递且具有高度容错特性的一致性算法）和PBFT（practical byzantine fault tolerant，拜占庭容错），而且支撑着这些内容的是Lamport（Paxos提出者），以及大量理论，PhD，论文，图灵奖！总之就是各种复杂的东西。

These above variations on the theme suffer from some pretty serious limitations, chief of which are,

they are administratively centralised, and/or/therefore
all participants are known.

以上提到的这些内容都有一些挺严重的局限性，主要有：

从管理上来说他们是中心化的，和/或/因此
所有的参与者都是已知的

In other words, the various Byzantine Generals Solutions assume that we know who the generals are, and famously that’s not usefully true in all cases.

换句话说，拜占庭将军问题的解决方案中我们知道谁是将军的假设，在以上提到的这些情况中并不适用。

Proof of Work

工作证明

Satoshi observed that any centralised component can be attacked, and likely will be (Nakamoto, Bitcoin: A Peer-to-Peer Electronic Cash System). This observation was accurate : MTB shut its blinded eCash down due to unexpected adult shopping, DigiCash was hamstrung by authorities, WebMoney was hit by a reputation attack, e-gold was brought down by the Feds, as was LibertyReserve. DigiGold and e-Bullion failed due to founder actions. The list is exhausting, and your chastened author was closer to some of these disasters than was comfortable. You should do your own research to eliminate blame, forgetfulness and other biases.

中本聪发现任何中心化的部件都能够被攻击，并会受到攻击（Nakamoto, Bitcoin: A Peer-to-Peer Electronic Cash System）。这个发现是准确的：MTB关停了eCash系统就是因为遭遇了意料之外的成人购物，DigiCash被政府当局干预致残，WebMoney遭到名誉上的攻击，e-gold和LibertyReserve被美联储扳倒。这个名单可以一直列下去，即使是作者本人也比预想中里这些灾难更近。你可以自己做一些调查来终结这些指责，以及其他一些偏见。

This was observed by several (see for example “The Mining Delusion”) but it was Bitcoin that presented a solution.

觉察到这个缺陷的人不止一个(例如参见"The Mining Delusion")，但是是比特币给出了解决方案。

Therefore, Satoshi argued, we need to eliminate the centralised vulnerability party (CVP, also known as a trusted third party or TTP). And by deduction, as we can’t trust voting by known parties, we must all share and prove the same data, and we must accept and relish easy entry and exit - psuedonymity.

因此，中本聪说，我们需要终结这种易受攻击的中心化机构（CVP，也称TTP）。并且，通过推测我们也知道并不能信任已知机构的投票系统，我们必须分享并验证这些数据是相同的，并且同时我们必须有容易接入的出入口，即匿名性。

Great stuff, but we still haven’t solved the double-spend problem, we’ve just moved it from a single place to a very much larger place :-(

这个东西很棒，但是仍不能让我们解决双重支付问题，我们只是将数据从单个地方挪到了一个大的多的地方。

The elegant and famous solution to this was proof of work (POW) or the nakamoto signature - a lottery based on hash-puzzle over a correct block of transactions. The cryptographic nature of the hash lottery selects one single miner at random, who produces the block. Coupling the costly hash search called mining with a reward and adding in some complicated game theory and probability is all designed to keep the miners on the straight and narrow. See comic.

一个优雅且著名的解决方案是工作证明（PoW）或者中本签名，一个基于用哈希数学问题来确认正确区块交易的彩票系统。这个彩票系统的密码学的本质随机选择一个制造区块的矿机。将被称为挖矿的昂贵的哈希搜索和奖励耦合，并加上一些繁琐的博弈论和概率问题，所有这些元素都被用来将系统设计得能够驱使矿机不停工作。详情见comic。

Bitcoin is a brilliant and elegant solution because it opens our thinking to the possibility of fully distributed applications, with money. But PoW burns up energy to the value that the market can bear, which amounts to a horrible tax on the entire value of the currency (as of time of writing, 4% on Bitcoin, and 11% on Ethereum, big ouch!) and as the Bitcoin chain moves to a fee base this means fees will bite hard, lifting Bitcoin out of reach of most people. High rewards and the rising price also resulted in economies of scale for mining, resulting inevitably in the concentration of miners. Although the system itself happily carries on, a perverse consequence of the censorship-resistant design is that the power of censorship now rests in the hands of about a dozen businesses, with most in one country that is not famous for resisting the urge to censor.

比特币是一个出色并且优雅的解决方案，因为它开阔了我们的思维，让我们想到了完全的去中心化式的应用（或货币）的可能性。但是PoW消耗了太多能量，多到市场无法承受，多到能够成为一项附着在货币上的吓人的税（比特币4%，以太坊11%！）。而随着使用比特币的费用越来越高，大多数人将无法使用它。高额回报和上升的价格同样也导致了大规模的挖矿，并不可避免的导致了矿机的中心化。虽然系统本身依旧毫无问题的运行，但这由能够抵御审查的设计方案完成的系统却落入了少数几个公司，而这几个公司几乎都坐落在一个以审查严格而闻名的国家。

Proof of Stake

权益证明

It was observed by someone (?) that we could simply replace the voting-with-CPU with voting-with-value in order to choose who makes the decision on (the next block of) double-spends. After all, the blockchain precisely establishes who owns what currency, and those who have more skin in the game are more likely to preserve the system, so it is an aligned bias. If uncomfortable to the small player, and somewhat offensive to democratic principles.

有人提出我们只需要将以CPU投票的模式改为以价值投票的模式来选择谁对（下个区块中）双重支付的问题做出决策。毕竟，区块链十分准确的建立在货币拥有的制度上，因此拥有更多货币的人更可能选择保护系统，因此这个偏向性是合理的。虽然可能让小额货币拥有者感到不适，也对民主制度有所冒犯。

So the theory goes. In practice, it has been criticised for (1) placing the power in the hands of those with most value, and also (2) for the “nothing at stake” problem which occurs when a bet on an alternate chain does not cost if it doesn’t survive.

这个理论也因为一下原因受到批判：（1）将权力交给最多价值拥有者手中；（2）“无需抵押”问题（the “nothing at stake” problem）：因为不需要付出任何代价所以会不断试图分叉（译者注：而对于矿机来说最优的策略是尽可能多的挖矿（制造区块），因此不论矿机挖的区块所属于的区块链最终是否获胜成为正确的那一条链，都能获得交易费。）

Simple proof of stake then does not seem to work. Let’s see what does work - let’s break down the problem.

简单的权益证明似乎并不奏效，究竟什么系统才奏效呢？让我们将问题逐步分解。

Let’s go to a Mining Centralisation Conference

让我们去进行一个挖矿中心化的会议

When we look at Bitcoin’s current state of a dozen or so well-known mostly Chinese miners, it is clear that they are all known, to us and to each other, and they can and do communicate. When we get to mining pools the size of today’s country-warming rigs, Bitcoin’s assumption of psuedonymity for miners becomes tenuous - just follow the electricity. Or go to any bitcoin scaling conference.

当我们观察到比特币矿工由中国垄断的现状时，很明显他们互相认识且能够相互沟通。而当面对今天大到能将国家变暖的挖矿规模情况下，我们知道比特币对矿工的匿名假设早已变得虚有其表了，因为我们只需要看电力消费或参与任意一个比特币会议就能知道到底是谁在挖矿。

And they could collude. So far, they have not chosen to do so, or, at least not obviously. As that’s a situation that hasn’t proven reliable historically, maybe incentives & ethics can be bolstered?

另一方面，他们可以相互勾结。至今他们并没有选择这么做，或者只是我们不知道。因为从历史上说这个情况还不能被证明是可靠的，也许动机与道德会被改进提高？

Only their incentives and their ethics keep the miners from colluding; it is the case that miners have returned the occasional fat-finger error, ones in which a trader has accidentally sent many BTC in fees instead of many satoshi, so ethics has some play here. On the other hand, some miners have mined empty blocks, even when the queue of delayed transactions or ‘mempool’ is exploding.

只有他们的动机和道德能够让他们不相互勾结；在这种情况下矿工又回到了胖手指（fat-finger）失误的时候，有的交易者会意外的将一定数量的比特币而不是satoshi单位设为交易费，所以道德在这里确实起着一定的作用。另一方面，一些矿机会挖出空的区块，即使有成堆的被延迟的交易排成长队等待被处理。

Not only is today’s miner concentration and cooperation unanticipated, the Bitcoin design was deliberately focussed on an alternate vision. Is miner integrity a sustainable future? Integrity is a cause that hasn’t proven reliable historically in for example banking, so maybe we should continue to bolster the incentives & ethics of blockchain?

今天的矿机中心化不仅是意料之外的，也不符合比特币最初的设想。矿工的正直会是一个可持续的未来吗？从历史角度来讲，正直并未被证明是可靠的，想想银行就知道，所以我们真的应该继续支持依靠动机和道德来支撑的区块链吗？

What if we work backwards and accept the fact that the miners can be concentrated? And that we know who they are? Or more controversially, miners should be concentrated and should be identified?

如果我们退一步接受矿工中心化的事实呢？并且我们知道他们是谁？或者更进一步地假设，矿工应该中心化并通过身份验证？

If we accept miner concentration and miner identification, we could simply appoint them. But appointing the Asic Generals just brings us back to the original centralised vulnerability situation.

如果我们接受矿工中心化和矿工身份验证，那就可以指定矿工。但是指定ASIC矿机，想过没有，这样的进程会将我们带回最开始的容易受攻击的中心化局面。

Delegated Proof of Stake - How to rule the Generals

DPOS-怎么管理将军

How then would we govern miners better? If we can appoint them, we can dismiss them as well, which brings us back to the Two Generals’ Problem.

那么我们究竟该如何更好的管理矿工呢？如果我们可以选出他们，我们同样可以解雇他们，这就将我们带回了两军悖论.

Given a new set of requirements forced on us by reality and experience, it is plausible to re-design the double spend system. Delegated Proof of Stake is just such a redesign, by @dantheman, using a combination of the tools above:

鉴于现实和经验强加给我们的新要求，重新设计双重支付系统是合理的。委托证明是由@dantheman重新设计的，使用以下工具：

1.A Producer (nee miner) is selected as decision maker to prevent double spends in one block. See Figure 4. Each block produced is rewarded by new currency (no change here).

1.一个生产者（nee miner）被选为决策者来避免一个区块中的双重支付，见图4，每次出块都能得到新的区块代币作为出块奖励；

2.Many producers are selected and given a round-robin rotation for a round of blocks, thus creating a competitive market within the round, ensuring overall reliability, and resolving forks. See Figure 5.

2.很多生产者被选择并被给予一个由一圈区块构成的罗宾环，从而在环内创造出一个存在竞争的市场，以保证可靠性并解决分叉，见图5；

3.The chain runs an open community poll to manage the producers, in which each member may vote according to proof of stake. Producers are both selected and dismissed in the same way - an auction for the next round. See Figure 6.

3.链上运行一个开放的社区通过投票来管理生产者，每个成员通过PoS原理进行投票。选择和解雇BP使用的是同一原理，即多轮拍卖式。见图6 。

The community then is required to govern their chain by

checking and agreeing with the transactions they perform,
voting the producers in and out based on their record and other pronouncements (e.g,. the producers are free to offer incentives such as revenue splitting) and
maintaining their stake or suffrage to the needs they choose.

然后社区需要通过以下行为来管理他们的链

检查并同意他们执行的交易，
根据他们的记录和其他声明（例如，生产者可以自由提供收入分配等激励措施）选出或者解雇生产者
维持他们的股权或选举权，以满足他们的选择。

Note that a more technical approach to attacks on DPOS can be found in DPOS Consensus Algorithm.

请注意，在DPOS共识算法中可以找到更具技术性的DPOS攻击方法。

A Political Economy?

政治化经济学？

In consensus terms, DPOS is stake-delegation over proof - the producers provide the proof over the blocks, and the community uses its stake to delegate the producers.

从共识的角度来说，DPoS是基于证据（proof）的代理权益（stake-delegation），即生产者向社区提供自己出块的证据，然后社区使用其权益来管理生产者。

In political terms, DPOS is similar to a two-layer representative democracy with landowner suffrage. Where,

从政治的角度来说，DPoS与基于土地拥有者投票的两层代理民主制度（two-layerrepresentative democracy with landowner suffrage）较为相似。在这种制度中：

representatives are those producers that are delegated by the community to decide the day to day questions (over double spend), and
suffrage, or the right to vote on representatives, is given to those who hold a certain form of property. This property might be a savings unit which represents a commitment to the community, and also a loss of liquidity; it resembles the historical landowner suffrage popular before more universal forms, as recognition of the wealth and commerce that the merchant class brought to society.

代表（representatives）即为那些由社区选举出的生产者，有生产者来处理那些日常的问题；
对拥有某种形式财产的人给予选举权或选举权。这个财产可能是一个储蓄单位，代表了对社区的承诺，也失去了流动性; 它类似于更普遍形式之前流行的历史土地所有者选举权，即对商人阶级为社会带来的财富和商业的承认。

The precise design of this mechanism - savings account, how many producers, how long a block, how long a round, what happens if?? - is obviously a deep and interesting question, and we’ll not go there today.

这种机制的准确设计——存款帐号，有多少出块者，多久出一个块，多长时间能形成一个循环，以及众多其他元素，显然是更深层次也是更有趣的问题，今天就不做讨论了。

How does it Perform?

它如何运作？

This mechanism has been shown to work in at least 2 large scale systems, Bitshares and Steem, and of course DPOS is to be used in EOS. It has also been chosen by Tezos, PeerPlays and Ark.io.

这个机制一直被不少具有一定规模的系统使用，如Bitshare，Steem，EOS，Tezos，PeerPlays，Ark.io等等。

While it is possible to criticise, so far the criticisms seem to be more at the level of flavour and comparison than weaknesses:

虽然可以对其批判，但迄今为止评论似乎更频繁的讨论其特点以及与其他机制比较，而不是其弱点：

1.Proof of stake is weak because of “nothing at stake” but this only matters when the stake is put to the direct question of a block. Delegation solves that - in DPOS the stake is put to the vote on the Producers, while the block is handled by a direct Producer round. Separation by delegation solves the “nothing at stake” problem.

1.PoS因为“无需抵押（nothing at stake）”问题而变得脆弱，但只有当权益（stake）被用来直接参与区块建设的时候这个问题才会出现。而DPoS则解决了这个问题，因为在DPoS中对区块的处理被交给了生产者，这种责任的分离解决了“无需抵押问题”；

1.There are possibilities for the producers to behave badly - they can censor transactions. But this is actually a flipped into benefit because badly behaved producers can also be voted out - that’s the point of delegation. Contrasting with Bitcoin’s current malpractice of the month of mining empty blocks, DPOS performs far better because it has a punishment mechanism.

2.有几种生产者会出现恶意行为的可能——他们可能会对交易进行干涉。但这事实上可以被转换为好吃，因为不合格的生产者会通过投票被解雇，这就是DPoS的意义。与比特币现在制造空区块的非法行为相比，DPoS的表现要强得多，因为它有一个惩罚机制；

3.There are possibilities to collude. Of course, collusion is also possible in other chains, as the above photo suggests. The question is really about which will perform better under collusion, and so far, our money’s on the organised governance because of punishment - skin in the game. The ad hoc or un-governed arrangement of pure PoW means that miners can’t be punished, even if they decide to mount their fabled 51% attack. In DPOS, once collusion is surfaced, it’s possible to adjust governance rules to deal with it be it at 51% or at 1%.

3.勾结依然存在可能性。当然，勾结在其他区块链也存在可能性。重要的是哪一种机制在勾结存在的情况下运行的更好，至今为止这我们的钱都得到了妥善的管理，因为惩罚机制的存在。而在缺乏管理的纯PoW中，不管矿工做什么都不会受到惩罚，即使是51%攻击。在DPoS中，只要勾结被发现了，不管是1%还是51%我们都可以通过调整管理方式来对付。

4.The purist bitcoiner will point out that by adding a governance layer, we’ve broken the trustless nature of the blockchain. Not so, as trustlessness rests on fallacies:

4.纯粹的比特币使用者会指出，添加管理层打破了区块链的无信任本质。然而并非如此，因为关于无信任有以下谬误：

a.Fallacy that there is no governance layer in other chains. In practice there are governance layers, but they are unwritten, denied, and inconstant or abused. In essence, we are replacing ad hoc (anarchic? captured?) governance with written, constitutional, formal and transparent governance. The question is not whether this adds governance, but whether explicit governance is better than the hidden intrigue.

b.Fallacy that the Bitcoin design means we don’t have to trust the miners. Consider the empty block syndrome again: we now have to trust the miners to produce decent blocks for the community, and they don’t, which exposes the gap between the mathematical nature of the blockchain and the human nature of ‘trust.’ The blockchain under PoW eliminates some trust but not all, and in this case, it destroys the possibility for the trust when it is needed.

a.其他区块链没有管理层的谬误。这并不正确，事实上是由管理层的，只不过没有被明写出来，被拒绝承认，以及不稳定。从本质上来说，我们使用有明确规章制度和透明的管理取代了专设的管理。问题不在于这么做是否加强了管理程度，而在一明确的管理是否比暗地里管理更好。

b.比特币的设计然我们不需要信任矿工。在考虑一下空区块的问题：我们现在必须相信矿工会生产有效的区块，而事实并非如此，这就将区块链的数学上本质和“信任”的人性本质分离开来。PoW下的区块链终结了一部分信任，但并不是所有，而在这种情况下，当我们真正需要信任时，我们却无法再建立信任。

DPOS comes with some advantages that are also worth stressing:

DPOS还有一些值得被提及的优点:

1.As it eliminates the hash mining, it is far cheaper than Proof of Work. In effect we are releasing the mining tax back to the community - as of writing 4% Bitcoin and 11% Ethereum.

1.由于其废除了哈希挖矿，DPoS会比PoW便宜得多。而收益的会是社区，因为我们会将这部分税归还给社区（如前面写到的比特币4%，以太坊11%）。

2.And we’re cleaning up the planet :-) DPOS is recommended for any blockchain with a heart ♡

2.DPoS更加环保。DPOS推荐使用任务有环保之心的区块链。

3.By adding a layer of formalised governance, we also set the scene for upgrades to the software. That is, users have a real mechanism to vote on a change, and producers have a real vote to follow. This mechanism pretty much makes the adversarial fork debate go away, and makes the chain fluid and dynamic - it can evolve quickly to suit evolving needs. I.e., the unsolved PoW nightmares of DAO and the blocksize debate are solved problems under DPOS.

3.通过添加正式的管理层，我们也能够有计划的为软件进行升级。这意味着，用户将会有机会对系统的改变做出投票，生产者根据投票结果行动。这个机制几乎使恶意分叉成为不可能，并使区块链充满活力，由于它可以随着不断变化的需求而进化。比如：仍未解决的PoW中DAO（decentralizedautonomous organization，去中心化自治组织）的梦魇，以及关于区块大小的争论都是在DPoS中得到解决的问题。

4.The formalised governance also works to give community the voice over their chain.

4.正式的管理机制也给了社区对改进区块链提出建议的机会。

5.DPOS works to channel the decision making into a high-performance and high-efficiency rig that allows truly massive throughput. Let’s leave performance to other posts, but it’s worth stating that Bitshares has tested at 1000X that which Bitcoin can achieve in its current form. That sort of performance creates a lot of headroom for forgiveness.

5.DPoS致力于使做决策变得高效，从而使系统真正能被大规模使用。让我们把关于性能的讨论留给其他文章，但这里值得一提的是，测试中Bitshare实现了1000倍于当下比特币的性能。这样的性能让人能够容忍很多其他的瑕疵。

Benefits are of course all sins to some, but they are the sorts of sins that businesses and individuals can forgive and treasure.

这些优点当然也意味着存在尚未被发现的缺点，但他们会是企业和个人能够容忍的缺点。

本文图片来源于英文原文

以下内容来自微信公共帐号“EOS技术爱好者”，搜索“EOSTechLover”即可订阅，翻译Gavin，校对Lochaiching。转载必须保留以上声明。仅授权原文转载。

本文原文链接为https://busy.org/@iang/seeking-consensus-on-consensus-dpos-or-delegated-proof-of-stake-and-the-two-generals-problem

"EOS技术爱好者"全程由EOShenzhen运营，喜欢我们请为我们投票：（EOShenzhen的投票账号：eoshenzhenio）！

了解更多关于EOShenzhen：

We are EOShenzhen

不同入口如何投票：
imToken
火币
 portal

关于我们更多联系：
Website：https://eoshenzhen.io

Steem：https://steemit.com/@eoshenzhen