Gavin Andresen做客巴比特 部分內容節錄

[yanlap]

1.crypto888
问一下Gavin和Blockstream的关系目前如何？ 是处于冷战期吗？当年你对BS的评价还是很高的： “If this amazing team offers you a job, you should take it,” tweeted Gavin Andresen, Chief Scientist, Bitcoin Foundation.    现在你对BS公司的看法如何？
What is your relationship with Blockstream now? Are you in a Cold War? Your evaluation on BS was pretty high “If this amazing team offers you a job, you should take it,” tweeted Gavin Andresen, Chief Scientist, Bitcoin Foundation.”  But now, what’s your opinion on BS?


I think everybody at Blockstream wants Bitcoin to succeed, and I respect and appreciate great work being done for Bitcoin by people at Blockstream.

We strongly disagree on priorities and timing; I think the risks of increasing the block size limit right away are very small. I see evidence of people and businesses getting frustrated by the limit and choosing to use something else (like Ethereum or a private blockchain); it is impossible to know for certain how dangerous that is for Bitcoin, but I believe it is more danger than the very small risk of simply increasing or eliminating the block size limit.

    我认为在Blockstream的每个人都想让比特币成功，我尊重并感激Blockstream的人为比特币做出的贡献。
    不过我们在一些事情的轻重缓急上有着很大的分歧。我认为立即提升区块上限带来的风险小之又小，同时我看到了个人和企业因为当下的上限对比特币望而却步，从而投向其他资产（比如以太坊和私链），虽然这对比特币带来的风险至今还未知，但是我认为这些风险一定是大于提升或者取消区块上限的风险的。


2.玛_雅
1、为何执意坚持Classic仅仅75%就硬分叉，从而3:1的算力比带来分裂币圈的风险。您之前答复是说因若95%可能控制5%而被控制，这个同意，但有个度。现75%是每四个人就有一个人不支持时，还强制进行硬分叉忽略高达1/4人的观点是否太偏激。感觉90%是否是更适合的度？
1. Why insist on hard fork at only 75%? You once explained that it is possible to be controlled by 5% if we set the threshold at 95%. I agree, but there should be some balance here. 75% means a high risk in splitting, isn’t it too aggressive? Is it better if we set it to 90%?

The experience of the last two consensus changes is that miners very quickly switch once consensus reaches 75% -- the last soft fork went from 75% support to well over 95% support in less than one week. So I’m very confident that miners will all upgrade once the 75% threshold is reached, and BIP109 gives them 28 days to do so. No miner wants to create blocks that will not be accepted by the network.

从上两次的共识分叉的经验来看，一旦共识达到75%，剩下的矿工都非常迅速的切到多数链上了。上一次的软分叉中，从75%到95%只花了一个星期。所以我相信一旦算力支持达到75%，剩下的所有矿工都会升级的，更不用说BIP109（Classic）给出了那些矿工28天的时间了。没有任何矿工愿意产出这个网络不接受的区块。

[yanlap]

2、怎么解决Classic路线图中，后期浮动上限，可能造成的超大区块，进而未来节点中心化的问题？
2. How to solve the potentially very large blocks problem Classic roadmap may cause, and furthur causing the centralization of nodes in the future?

Andreas Antonopoulos gave a great talk recently about how people repeatedly predicted that the Internet would fail to scale. Smart engineers proved them wrong again and again, and are still busy proving them wrong today (which is why I enjoy streaming video over my internet connection just about every night).

I began my career working on 3D graphics software, and saw how quickly we went from being able to draw very simple scenes to today’s technology that is able to render hundreds of millions of triangles per second.

Processing financial transactions is much easier than simulating reality. Bitcoin can easily scale to handle thousands of transactions per second, even on existing computers and internet connections, and even without the software optimizations that are already planned.

    Andreas Antonopoulos最近有个关于人们曾经如何多次预测互联网没法扩容的演讲。聪明的工程师们一次又一次的证明这些人的说法是错误的。（这就是为什么我每晚看网络视频都看的很开心的原因）
    我的第一份工作是写3D作图软件，我亲眼目睹这个行业从一开始只能画很简单的场景到今天的每秒渲染上亿个三角的迅速变迁。
     处理金融交易比起仿真现实可容易多了，在今天的电脑和网络条件下，就算不实施未来那些软件优化，比特币都可以轻松的扩容至每秒上千个交易。

3、请问您为何不支持中本聪早期最早提出的RBF提案，甚至打算在Classic中完全删除它。
3. Why do you not support the proposal of RBF by Satoshi, and even plan to remove it in Classic completely?

Replace-by-fee should be supported by most of the wallets people are using before it is supported by the network. Implementing replace-by-fee is very hard for a wallet, especially multi-signature and hardware wallets that might not be connected to the network all of the time.

When lots of wallet developers start saying that replace-by-fee is a great idea, then supporting it at the network level makes sense. Not before.

    只有在当大多数钱包和大众都开始用了RBF的时候，才是该讨论比特币网络支持RBF的时候，因为由于multi-signature和硬件钱包并不能随时保持网络联机的原因，在这些钱包上面实现RBF是非常困难的事情。
    只有当许多钱包开发者开始认为RBF是个很好的主意后，才是在比特币网络上布局RBF的时候，而不是反之。

[yanlap]

4、对于隔离验证软分叉、侧链技术、闪电网络技术，您是支持还是反对，请简述下理由，谢谢。
4. Your opinion on soft fork SegWit, sidechain, lighnting network. Are you for or against, please give brief reasons. Thanks.

The best way to be successful is to let people try lots of different things. Many of them won’t be successful, but that is not a problem as long as some of them are successful.

I think segregated witness is a great idea. It would be a little bit simpler as a hard fork instead of a soft fork (it would be better to put the merkle root for the witness data into the merkle root in the block header instead of putting it inside a transaction), but overall the design is good.

I think sidechains are a good idea, but the main problem is finding a good way to keep them secure. I think the best uses of sidechains will be to publish “write-only” public information involving bitcoin. For example, I would like to see a Bitcoin exchange experiment with putting all bids and asks and trades on a sidechain that they secure themselves, so their customers can verify that their orders are being carried out faithfully and nobody at the exchanges is “front-running” them.

    只有让大众亲自去尝试一件事物才是验证一件事物能不能成功的最好方式。也许很多事物会失败，不过只要有一部分能成功，那就不是问题。
    我认为SegWig是一个很棒的主意，并且使用硬分叉会比使用软分叉要简单一些。（因为把给witness data的merkle root放入block header会比将之放入交易中更好。）但是总体设计是很好的。
    我认为sidechain是很好的主意，但是主要问题在于如何确保他们的安全性。我认为最好的侧链模式应该是只发布在比特币中的“write-only”的公共信息。比如：我希望看到有交易所去试着将买单卖单都放在侧链上，并且在侧链上进行交易，因为侧链能够保证这些交易的安全性。这样一来客户可以验证每单交易是否都如实执行，没有任何人能做任何猫腻。

5、脑钱包脑口令工具，您之前发文《千万别用脑钱包!你很可能失去你的比特币!》（可能译者有点夸大或误解）http://8btc.com/thread-2783-1-1.html，请评价下新文章《比特币脑钱包，你应该用！》http://www.8btc.com/bitcoin-brainwallet《左脑右脑一个完美组合，双脑脑钱包诞生》http://www.8btc.com/brainwallet-2
能否说下您对脑钱包脑口令工具的最新观点？若解决了新用户难使用长且复杂脑口令的安全问题后，是否其应该是不错的工具？
5. Can you share your latest opinion on Brainwallet? It is hard for new users to use long and complex secure passphrase, but is it a good tool if it solves this problem?

We are very, very bad at creating long and complex passphrases that are random enough to be secure. And we are very good at forgetting things.

We are much better at keeping physical items secure, so I am much more excited about hardware wallets and paper wallets than I am about brain wallets. I don’t trust

[yanlap]

myself to keep any bitcoin in a brain wallet, and do not recommend them for anybody else, either.

    我们人类在创建冗长且复杂的密码方面其实非常差劲，经常会出现随机性不够强，不够安全的密码，且我们记忆也不是那么好。
    我们更擅长于安全保存好一个实物，所以比起脑钱包我对硬件钱包和纸钱包更感兴趣。我不相信我自己能够在脑钱包里保存好比特币，也不推荐其他人这样做。


3.比特吹
Gavin 你现在手上还有比特币吗， 在MIT的主要工作是啥， 是否有FBI的人来追查你过。 你认为SHA256可能在什么时候被淘汰， 现在看来似乎已经有点不安全了。
Gavin, do you have bitcoins now? What is your major job in MIT? Has FBI ever investigated on you? When do you think SHA256 might be outdated, it seems like it has been a bit unsafe?

Yes, a majority of my own person wealth is still in bitcoins -- more than a financial advisor would say is wise.

My job at MIT is to make Bitcoin better, in whatever way I think best. That is the same major job I had at the Bitcoin Foundation. Sometimes I think the best way to make Bitcoin better is to write some code, sometimes to write a blog post about what I see happening in the Bitcoin world, and sometimes to travel and speak to people.

The FBI (or any other law enforcement agency) has never investigated me, as far as I know. The closest thing to an investigation was an afternoon I spent at the Securities and Exchange Commission in Washington, DC. They were interested in how I and the other Bitcoin developers created the software and how much control we have over whether or not people choose to run the software that we create.

“Safe or unsafe” is not the way to think about cryptographic algorithms like SHA256. They do not suddenly go from being 100% secure for everything to completely insecure for everything. I think SHA256 will be safe enough to use in the all ways that Bitcoin is using it for at least ten years, and will be good enough to be used as the proof-of-work algorithm forever.

It is much more likely that ECDSA, the signature algorithm Bitcoin is using today, will start to become less safe in the next ten or twenty years, but developer are already working on replacements (like Schnorr signatures).

    有的，我绝大多数个人财产都在比特币里——远超过理财师的建议。
    我在MIT的主要工作是让比特币以我认为最好的的方式变得更好。这和我在Bitcoin Foundation中的工作是一样的。有时候我认为写代码能让比特币变得更好，有时候我认为是将我的在比特币世界中的所见所闻写成博客，有时候是旅行并和人们交谈。
    我从来没有被FBI（或者其他执法机关）调查过。只有一次经历可以算作类似调查，那是在华盛顿DC的证券交易委员会的一个下午。他们问我和其他的比特币开发者是如何开发出比特币软件的，并且问我们对于人们是否运行比特币软件有多少的控制力。

[yanlap]

 “安不安全”并不是描述SHA256密码算法的最好方式，因为他们不会一下从100%对所有东西都安全变成完完全全对所有东西都不安全。我认为按照SHA256目前在比特币里的各种使用场景来看，这种算法至少在十年内都还是足够安全的，甚至可以永远用它作为工作量证明的算法。我认为SHA256至少在十年以内对比特币所有的用途都足够安全，并且它用在PoW算法上永远都足够好。
    应该这么说，比特币今天的签名算法ECDSA会在十到二十年后开始渐渐变得不安全，不过现在开发者们已经开始着手于它的替代品了（比如Schonrr签名）。

4.三胖很棒
久仰久仰！ 只问一句，Gavin目前有为什么公司服务吗？或者说从谁哪里领的工资 ？
It’s a pleasure to meet you. I only have one question. Which company are you serving?  or where do you get your salary?


The Media Lab at MIT (Massachusetts Institute of Technology) pays my salary; I don’t receive regular payments from anybody else.

I have received small amounts of stock options in exchange for being a techical advisor to several Bitcoin companies (Coinbase, BitPay, Bloq, Xapo, Digital Currency Group, CoinLab, TruCoin, Chain) which might be worth money some day if one or more of those companies do very well. I make it very clear to these companies that my priority is to make Bitcoin better, and my goal in being an advisor to them is to learn more about the problems they face as they try to bring Bitcoin to more of their customers.

And I am sometimes (once or twice a year) paid to speak at events.

    我的工资是MIT的Media Lab发给我的，除此之外，我没有其他的持续性收入。
    作为技术顾问，我会时不时收到比特币公司（Coinbase, BitPay, Bloq, Xapo, Digital Currency Group, CoinLab, TruCoin, Chain）的小额股票期权，如果这些公司发展非常好的话，这些期权也许有一天也会很值钱。我也很明确得告诉这些公司，我的首要任务是让比特币变得更好，当他们的技术顾问的目的也是为了了解他们公司在将比特币推给客户时候面对的问题。
    另外，我还有时会有收到一些演讲出场费（一年一到两次吧）。

5.洒脱喜
您对闪电网络（LN） 这种 layer 2 的协议是怎么看的，它的实现复杂么？  需要通过硬分叉才能实施吧？
Would you mind share your opinion on lightning network? Is it complicated to implement? Does it need hard fork?


Lightning does not need a hard fork.

It is not too hard to implement at the Bitcoin protocol level, but it is much more complicated to create a wallet capable of handling Lightning network payments properly.

I think Lightning is very exciting for new kinds of payments (like machine-to-machine payments that might happen hundreds of times per minute), but I am skeptical that it will be used for the kinds of payments that are common on the Bitcoin network today, because they will be more complicated both for wallet software and for people to understand.

    闪电网络不需要硬分叉。
    在比特币协议层实现它并不复杂，难的是建立能够正确处理闪电网络的钱包。
    我认为闪电网络在一些新兴支付方面非常棒（如：机器对机器之间的每分钟上百次的支付），但是我对它能否用于今天的比特币网络的支付持怀疑态度，因为它即复杂化了钱包，又让大众难以理解。

[yanlap]

完整内容版本：

https://docs.google.com/document/d/1p3DWMfeGHBL6pk4Hu0efgQWGsUAdFNK6zLHubn5chJo/edit?pref=2&pli=1

[jyaken]

谁知道扩容的事有进展吗？

[Aimee1993]

以太小王子被请到中国来了

[a7594li]

比特币最近价格上来了, 这几年的确得到了大众的认可!

[hao567tao]

谢谢

[hao567tao]

同问

[hao567tao]

小王子哈哈

[hao567tao]

大势所趋