Hard Forks

The crypto-currency Ethereum completed a hard fork on Wednesday. The Ethereum core developers, after getting a vote of support from the Ethereum community, hard forked Ethereum to “get back” the roughly $40mm of Ethereum that was taken in the hack of The DAO.

Hard forks are a bit of a lightning rod issue in the blockchain sector. The Bitcoin community has been debating the idea of doing a hard fork to increase the block size for well over a year. It seems that most of the Bitcoin core developers are against a hard fork and see it as risky. Bitcoin did have an accidental hard fork back in 2013, but that was dealt with quickly and confidence in the Bitcoin blockchain was restored.

I believe that hard forks are an inevitable occurrence in the blockchain sector. There have been, and will continue to be, issues that crop up that are best solved with hard forks. I do not think they will be common and I do not think they should be common. But they are an important tool in the toolset that core developers have to move these protocols forward. And so I see the successful Ethereum hard fork this week as an important milestone for public blockchains.

I like what Cornell CS Professor Emin Gün Sirer said about the Ethereum hard fork:

It’s a point of strength to be able to adapt to that change, to be able to respond to it, to be able to do it in an orderly fashion. Ethereum just demonstrated this. I think this is a rite of passage for ethereum.

In my mind, adaptability is more important than immutability.

And to some extent, that is what is now at the center of the Bitcoin vs Ethereum competition for the hearts and mind of developers. I believe the Bitcoin core developers have more or less landed on immutability and Ethereum core developers are very much into adaptability. It may be that it is useful to have two significant, liquid, and highly capitalized public blockchains, one that is immutable (think of gold) and one that is adatpable (think of the dollar).

There was a time when I was a Bitcoin absolutist. That changed a while ago and I now believe that we are going to have multiple blockchains, multiple currencies, and a ton of app tokens, some with their own blockchains, some built on top of Bitcoin or Ethereum.

It’s a very interesting time in the public blockchain sector right now. Stuff is happening. Lot’s of stuff.

#blockchain#Uncategorized

Comments (Archived):

  1. LIAD

    “Reports of Your Retirement Have Been Been Greatly Exaggerated”

    1. fredwilson

      LOL

  2. JimHirshfield

    “Bitcoin core developers have more or less landed on immutability and Ethereum core developers are very much into adaptability.”That’s quite a mouthful. But it has teeth. Tine will tell.

    1. Eric Satz

      Very punny

    2. Twain Twain

      Ethereum’s USP from Bitcoin is that it is Turing-complete. We can Google what Turing-complete means for mutability and adaptability.There’s an interesting Reddit thread on the maths:* https://www.reddit.com/r/et

      1. Soci

        thats a bit misleading ain’t it? any turing complete language can emulate any other language, including languages that arn’t turing complete, which in turn kinda invalidates alla that.You can decide to shoot yourself in the foot, or make something vastly more simple and safer

    3. Richard

      Based on the headline I thought this was going to be Fred’s take on the DJT speech

      1. fredwilson

        Fuck no. I didn’t watch it. I didn’t read about it. I am blissfully ignorant of what went down in Cleveland this past week

        1. Girish Mehta

          Deliberately ? (didn’t watch / read ). Or it just happened that way ?

          1. fredwilson

            Very deliberately. The antidote to an attention whore is to ignore them

          2. Girish Mehta

            OK.

        2. ShanaC

          twas interesting, in good and bad ways.What do you think of Gary Johnson – he’s catching up in the polls

        3. bernardlunn

          My wife has same policy, to just ignore it, which is wise. I must admit to some fascinatiion it was so crazy.

    4. ShanaC

      We’ll see. Wsj is reporting that ethereum developers have a goal of trying to stay as immutable as possible

  3. awaldstein

    Thanks–my education continues.I do seriously wonder if there are going to be so many different touch points to the market isn’t it overdue time that the sector thought and acted a bit on interfaces and communications that are just easier to use and clearer to understand.I’ve always worked on the premise that you need to address both the needs of the developers and that of the market itself. These are different hearts to be won.

    1. fredwilson

      Try steemit.com. I think they have done a better job of putting the blockchain stuff brhind the scenes and just making something people can and will use.You won’t find the links all that great. But you will get my point

      1. awaldstein

        Thanks–not kvetching btw, just noticing.Will do so.

  4. Eric Satz

    Adaptability to a little chaos goes a long way

  5. Tom Labus

    Hard forks act as good management which is required for both growth and stability

    1. jason wright

      if a blockchain can be hard forked it’s not decentralised.

      1. William Mougayar

        It was a community decision. The decentralized community voted.

        1. jason wright

          48% of Brits voted to stay in the EU.

          1. pointsnfigures

            They lost. Around the same number voted against Obama. They lost.

          2. jason wright

            a society doesn’t function well when almost half of its inhabitants are being dragged kicking and screaming to a place they don’t want to go to. the winner takes all approach is a fail in my book. it’s not the way to build a civilised world.

          3. William Mougayar

            true, and i wished voting for something like that required a higher majority, like 2/3 instead of 1/2 only.

        2. awaldstein

          Interesting….Can you describe the size of this community and how a vote or decision was made.Interested as a builder of community and it is freakin hard to get decisions from large communities of any sorts.

          1. William Mougayar

            It gets technical a bit. You vote with your wallet basically. You send 0 ETH to an address for Yes, or to another for No. The size of your balance is a weighing factor, i.e. if you have 100$ your vote counts 100x if you had $1. It’s not perfect, but it worked. here’s the link http://carbonvote.com/

          2. Jess Bachman

            The size of your balance is a weighing factor, i.e. if you have 100$ your vote counts 100x if you had $1. Wow. Crypto-capitalism at it’s finest.

          3. William Mougayar

            It is what it is. still way more decentralized than otherwise.

          4. Lawrence Brass

            Keep pushing blocks for the pyramid. Thchak! :-/

          5. Jess Bachman

            “It is often safer to be in blockchains than to be free.”- Franz Crypto-Kafka

          6. Girish Mehta

            “Man is born free, and everywhere he is in (block)chains” – H/T Jean-Jacques Rousseau

          7. Lawrence Brass

            “There is no honour in this blockchain”- Sgt. Shamar Crypto-Thomas

          8. Girish Mehta

            “We forge the (block)chains we wear in life “- H/T Charles Dickens.

          9. awaldstein

            So interesting… not democratic in the least and for me at least the only open community that would embrace this.Pure bastion of capitalism in a pure form.

          10. William Mougayar

            that’s not how they see it, i.e. the decentralization aspect is real. there are economic parameters that come into play, of course, but truth is that ownership is very distributed, so voting follows that same distribution. no single entity owns a large enough amount to sway it either way.

          11. awaldstein

            I actually like this is this instance but this is not a democratic process in any sense of the word and that is what I was querying.It is just like we all get the right to vote by being citizens but the more wealthy you are the more the vote counts.

          12. William Mougayar

            Let’s compare, and stay within a financial context. – do you get to vote to decide if your bank can charge you such and such fees, or on whether Visa wants to change their policy, or pick any big company you deal with, where you are a customer and they don’t ask you anything.- this is more like voting as a shareholder. if you own a stock, each share equals one vote (more or less), so that is a more equivalent process, although it is more spread out and less regulated within the crypto world- finally, this is about having a stake or not. if you have a stake (ownership), then you vote. if you don’t, you’re not a participant.all and all, cryptocurrency governance is not a bad alternative to top down, closely held decision-making. it’s not perfect, but it is more decentralized and accessible to the masses.

          13. awaldstein

            Nicely said.I get this completely.So this is not a community but a network of financial stakeholders, with power weighted by the level of your stake.All good.So the big question we should talk about sometime is what in actuality does this have to do with democratizing finance?Doesn’t have to be to be valuable and interesting. But doesn’t seem to at least by the direction of this discussion.

        3. SubstrateUndertow

          What voting mechanisms would make that less than fictionally feasible at scale ?

      2. Dan McArdle

        That’s just flat out false. PoW blockchains are mathematically subject to simple-majority rules. And a majority of independent actors all agreeing about something is not at all the same as a single entity or small group forcing a change on a much larger group. This idea that hard forks break decentralization is flat out wrong on both “common sense” and mathematical levels.

        1. jason wright

          i’m talking about vested interest politics. 51% can be a vested interest. look at the power of the few Chinese bitcoin miners.

          1. Sebastian Wain

            Indeed you can launch an attack with ~33% with selfish mining.

          2. Dan McArdle

            There’s probably a reason why this hasn’t happened in the wild yet…specifically, that miners realize they shouldn’t act in a way that undermines the integrity of the system as a whole, despite their short-term earnings, because it would be a bigger hit to their long-term earnings.

          3. Dan McArdle

            In theory, alignment should happen around the value of the chain. That is, if the market told the top 3 miners that their actions were destroying the value of Bitcoin, they’d change course because it’s in their economic interests to do so. That alignment defines the chain, whether you as an individual agree with it or call it a “vested interest”. Things only fall apart if you assume powerful actors who are not economically motivated (which is another class of discussion).

          4. SubstrateUndertow

            Isn’t that a very time lagged and psychological vague set of feedback loops on which to base any kind of collectively systemic stability/faith ?

      3. Tom Labus

        They should have let the cash go?

        1. jason wright

          banks and poor decisions and the moral hazard of a bailout. a hard fork is like a bailout.core team and weak code and hard fork. where’s the pain that will sharpen minds and modify behaviour?with a billion dollar market cap has Ethereum become too big to be allowed to fail, and is this hard fork creating an unhealthy precedent that will actually hurt the blockchain sector in the future?

          1. Lawrence Brass

            This is handcrafted human technology, what else would you expect? The public has the wrong idea that anything built on top of an uncorruptible ledger is by extension uncorruptible or right, and it’s not. You can write anything on it, not necessarily representing the intentions of the majority of the participants, specially when the one writing on it is an imperfect human being or his code.Vulcans only write perfect code.

          2. Tom Labus

            It’s their own cash

      4. Sebastian Wain

        You can say the same to improvements (BIPS in Bitcoin) when the last word is very centralized based on core developer decisions.

  6. jason wright

    blockchain – decentralised trustless counterparty platform.forkchain – something else.

    1. Marc Morin

      This is the core conclusion. Trustless meaning you can’t trust anyone is the key value.

      1. jason wright

        not needing to trust is possibly the greatest innovation and attraction of blockchain.a hard fork is a subjective intervention.

    2. William Mougayar

      Part of the same evolution.

      1. SubstrateUndertow

        “Evolution” is an emergent affair where the inevitable/repeatable complexity failures/errors are swamped out by massive functional redundancies.On its face that strategy does not seem to be applicable to financial exchange mechanisms at least not in its decentralized incarnations ?

  7. William Mougayar

    I agree with everything you said there. I would add that the month-long process that the Ethereum community undertook was key in letting them arrive at that decision. It had some ups and downs, but it resulted in a good outcome.As you said previously, this was the equivalent of an election. And in the course of 30 days roughly, Ethereum had the luxury of running an open debate, testing several options, voting on the choice, and implementing it. That was a commendable series of achievements that displayed its adaptability in more ways than one. Good thing that Vitalik and Emin were the sage voices that shepherd this process, and they were supported by the Ethereum community who responded and acted.

    1. jason wright

      but it has the feeling of the old ways of an institution like the Federal Reserve. i thought blockchain tech was all about eliminating policy and tweaks that favour vested interest.

      1. William Mougayar

        This is early days, so it’s ok to adjust when the blockchain isn’t even a year old. I think of it more as a tune-up than a policy intervention. And as Vitalik observed, we may not get the luxury of doing it again. We were lucky this time that the damage could be undone, but we may not be lucky again in the future.

        1. jason wright

          a blockchain with a billion dollar market cap will always be vulnerable to the certainty that there will always be more smart people outside the core team than inside, and therefore further exploits are inevitable.

          1. William Mougayar

            There will always be detractors of smart contracts and Turing completeness, but you can write bad and buggy software on anything. Only difference is that with smart contracts and blockchains, these are relatively powerful programs with money involved in the middle, so the outcome of messing-up is more visible. We just have to learn to be responsible about it.

          2. ZekeV

            The problem with Solidity is not that you can write buggy software. It is that, as a matter of computer scientific theory, any program written in a Turing-complete language might halt. And there is no way to know that by examining the code. That is an as-yet unsolved problem, in theory.This makes it risky to rely on a smart contract written in a Turing-complete language as the sole decider of how to allocate a high-value payout. That doesn’t mean there aren’t appropriate uses for Solidity, mind you. I just wouldn’t use it to contract for any significant financial value.

          3. William Mougayar

            btw- Ethereum is “quasi Turing complete” meaning that the programs don’t run indefinitely til they halt. But if you write a good smart contract, the outcomes are very interesting.

          4. ZekeV

            OK, so are you saying that eth contracts written in solidity do not suffer from the halting problem?

          5. William Mougayar

            that’s correct. there are inherent limits (implemented with the gas parameter) to how long they will run, so they can halt gracefully instead of running indefinitely.

          6. ZekeV

            I don’t think that solves the halting problem. It stops a transaction from tying up all eth nodes, b/c they stop running it when the contract lacks gas. The contract has still halted, however, and what happens to the parties who were relying on that contract to produce an outcome?

          7. p-air

            Oh yeah that, and smart contracts can never be deleted. How’s that for efficiency and scaling?

          8. Twain Twain

            I believe in smart contracts and in publicly distributed protocols.It’s the Turing-complete aspect wrt immutability & reversibility that mathematicians need to solve.The hard fork’s a necessary temporary solution but the underlying maths is what needs improving too.

          9. SubstrateUndertow

            “We just have to learn to be responsible about it.”What does that exactly mean in practice at scale?If at scale failures are potentially catastrophic dosen’t one need to avoid methods that are not provably determinant/patchable ?

          10. SubstrateUndertow

            Great read flying muchly over my simpleton head-space !I vaguely gather he is making the case that Turing-complete imperative programming language structures inherently exhibit a-priori unknowable/emergent indeterminacy-effects that are non-starters for mission critical financial contracts where as fictional DSL languages can possibly supply a formal a-priori provable determinacy about their outcome ranges.That feels coherent to me and my ignorant little bio-analog metaphoric obsession where I’d like to believe that an APIs-driven global-algorithmic-jazz-symphony of Turing-complete ecosystem agentscould =a possible substrate for an emergence brave new world of social/economic living systems evolution.color me crazy 🙂

          11. Twain Twain

            LOL, crazy can be good!It’s not even Turing-completeness that systems should be aiming for.It’s DaVinci-coherency.But that’s a whole other quantum level of innovation.

          12. Lawrence Brass

            I actually like your bio-analog metaphoric obsession, specially after I began to understand how to decode it. Awareness of self craziness is not crazy, it is powerful. Maybe you are a Zerg. 🙂

          13. ShanaC

            what is a zerg?

          14. Lawrence Brass

            In the context of starcraft, the game, it is a species or race you can choose to play with. I won’t try to define it further to avoid offending any fan, but there is a very detailed description at wikipedia. I used to play Terran.https://en.m.wikipedia.org/

        2. ZekeV

          A lot of smart people such as yourself and Brian Armstrong seem to view bitcoin and ethereum as software products. I would prefer for any protocol changes to be actual long-term improvements to the protocol, and not “bills of attainder” to undo undesirable transactions or frustrate a particular program, however malicious. But you know, this was fair game, and I’m not one of the folks who thinks any hard fork is heresy.

          1. William Mougayar

            Again, the “system” properties of Ethereum are preserved. I see this as a slight course correction. I think it would have been worse to let the thief get away with $50M. Then it would set a precedent for others.Neither solution was perfect, but the hard fork seemed to be the more prudent choice.

      2. fredwilson

        Meet the new boss, same as the old boss

    2. ZekeV

      It may have been silly to fork ETH to undo one exploit, of the many exploits that will likely be made of this protocol. At the same time it sets a precedent, and is a valuable lesson that any bitcoin clone is susceptible to a protocol change. It is dangerous for companies and speculators to have a strong vested interest in there not being a hard fork. It *will* happen, like it or not.

      1. William Mougayar

        Well, the hard fork didn’t hurt anybody really except the thief. So I’m not sure I see where the “dangerous” part is. The exploit was on the DAO, not Ethereum proper. The lessons here is that mart contracts in general will become stronger and more resilient.

        1. ZekeV

          Well let’s say you are party to a smart contract. Maybe a hedging contract involving credit default by a big bank, payable in eth. The reference event occurs, and you pocket a huge sum of eth. The public is outraged at this “fraud” or maybe there is a court decision, and this convinces the developers to roll back the blockchain to claw back your ill-gotten gains. Except your company was actually relying on this hedging contract and now you are broke.That is the danger, that a smart contract on ethereum is not really going to be honored. In the recent hard fork, the dishonoring is happening due to political pressure brought to bear on a group of young men who all have a vested interest not only in ethereum but in the short term value of The DAO. The point of the theft was to destroy The DAO, in a manner consistent with the explicit written policies announced by ethereum developers (immutability of eth smart contracts), as well as the anarchic philosophy of bitcoin. It is now clear that the idea of smart contract immutability is going to require a more change-resistant protocol than ethereum.

          1. William Mougayar

            I’m not sure I would interpret it this way.If you have a successful hedge contract and you make a profit without doing anything illegal, good for you. No one is going to roll back that contract. In the case of the DAO, aside that there were some Ethereum developers that also had stakes in the DAO, that wasn’t the driving factor.

          2. ZekeV

            The DAO “thief” arguably didn’t do anything illegal either. He did something that a lot of people found objectionable. I do not doubt the integrity of the ethereum developers. However, to have a stake in a particular smart contract, as well as ethereum the protocol, is clearly a conflict of interest. And an unavoidable one in the world of ethereum, unless the developers decide to avoid this conflict voluntarily.The entire point of smart contracts is that they are enforced in code, not in the courts, yes? Thus whether or not a particular contract involves fraud, the protocol absolutely should not be changed to undo the contract (this is my personal view anyway). The party who is defrauded has legal remedies, and if the loss is great enough, should pursue them conventionally — ie, go get a judgment and try to enforce it on the loser. Unfortunately this is not going to work where a contract is open to anyone joining. Thus if you have a lot invested in the smart contract, maybe you also should know exactly whom you are dealing with in physical reality, and have some larger legal framework with the “smart” portion of the contract handling only a portion of the transaction… Sorry this is a bit long-winded for a comment.

          3. William Mougayar

            OK, but if that hacker was an ethical hacker and wanted to divert that money to use it according to the DAO intent, then your argument holds; but in this case he/she/they didn’t exhibit any of that intent. They just wanted to keep the money, destroy the DAO, hurt Ethereum and maybe they made money already by shorting it.

          4. ZekeV

            I don’t think we have to inquire as to the cracker’s ethics or intent. And the entire point of smart contracts is that we do not have to debate such things, no?

          5. William Mougayar

            Good point, but when something of that scope happens, then you have to use some human judgement. As Vitalik pointed out, there were known weaknesses prior and some users lost some money here and there, and that was accepted of course. This is a really good read:https://blog.ethereum.org/2

          6. Lawrence Brass

            Enjoyed you and Zeke’s ping pong.From the perspective of my ignorance of the implementation details, it seems that people have assumed that the immutability of the underlying merkle trees in blockchains somewhat magically permeates upwards, wrongly giving immutable or uncorruptible status to every crazy thing they are doing at that upper level.Transactional behavior enforced by the actual implementation of Ethereum’s VM and the languages at its core are the top priority in my opinion, instead of looking for formal validation proofs. Stronger foundations will bring stronger constructs.After all, it is not magic, it is code.

          7. William Mougayar

            Right. and we’re fine tuning it. nothing is perfect from day 1…or year 1.

          8. Twain Twain

            Exactly this, Lawrence. The issue of immutability within Turing-complete systems:* https://www.reddit.com/r/et…Albert Wenger wrote: “And there is huge value in any system from ultimate reversibility. In fact we should be deeply careful with anything that is not reversible.”Turing-complete lends itself to irreversibility.

          9. p-air

            Once again, we must remind William what Stephan Tual parroted for months leading up to the deployment of TheDAO, which can also be found in the “Explanation of Terms and Disclaimer”:”The terms of The DAO Creation are set forth in the smart contract code existing on the Ethereum blockchain at 0xbb9bc244d798123fde783fcc1c72d3bb8c189413. Nothing in this explanation of terms or in any other document or communication may modify or add any additional obligations or guarantees beyond those set forth in The DAO’s code. Any and all explanatory terms or descriptions are merely offered for educational purposes and do not supercede or modify the express terms of The DAO’s code set forth on the blockchain; to the extent you believe there to be any conflict or discrepancy between the descriptions offered here and the functionality of The DAO’s code at 0xbb9bc244d798123fde783fcc1c72d3bb8c189413, The DAO’s code controls and sets forth all terms of The DAO Creation.”At no time do I recall William or anyone else from the Ethereum Foundation, Slock.it, or other DAO fanboy saying, “but in case of catastrophic failure or unintended consequences, the Ethereum Foundation should intervene”. Bank bailouts by any other name, remain bank bailouts 😉 And speaking of transparency, where can we see who all the holders of DAO tokens (and their percentage holdings) were? Hmmm…

          10. William Mougayar

            Not sure if you read my DAO analysis, but here is the link. http://startupmanagement.or…In a nutshell, I had been vocal in private circles against several aspects of the DAO, early on. So, this means that all their disclaimers didn’t hold water to me.

    3. JamesHRH

      If you put a customer hat on and read this post, the only potential reaction is:Huh?orWTF?You solved losing $40M by doing a hard fork? Don’t even bother trying to say that in Normal.

  8. wowjeff

    So basically you wrote all of this without learning the basics about databases?You can have a distributed database that is not a blockchain. You guys are fucking idiots.

    1. fredwilson

      And how do you really feel 😉

    2. SubstrateUndertow

      I’m OK with you calling me an idiot cause I’m the resident know-nothing peanut-gallery idiot around here.But this resident idiot has been around here long enough to notice that most of the others here are pretty informed/smart cookies who are openly sharing ideas without prejudice/attitude.You may be seeing the world as a reflection of yourself ?But if like this idiot you stick around here you too will be amazed at how much learning/perspective this community will provide to you 🙂

      1. brian_aaargstrong

        No, they are people that don’t realize they are simply rebuilding old systems and ADDING inefficiencies instead of REMOVING inefficiencies. Not to mention Coinbase is likely approaching running out of money, and is scrambling at ether to generate revenue, because they have been such a shitty company handling Bitcoin that people are going elsewhere.

  9. Dan McArdle

    I think you’re spot on re the different approaches Bitcoin and Ethereum are taking, and have noticed that the two communities attract different types of people, with different social values. Fascinating to see this reflected directly in the technology.On the poly-chain stuff – there may be many chains/coins/tokens, but most of the value will likely remain in the top handful; as a domain highly subject to network effects, it seems this should continue to follow something of a power law.

  10. ZekeV

    Users are free to continue running pre-fork ETH “classic” protocol. Bitcoin, or ethereum, absolutism does not make sense. It is one of the ways in which the p2p protocol market appears to be different from previous internet and web startup dynamics. In bitcoin we still have network effects, but I think in a non-exclusive way. It does not appear to be a winner-take-all market. There are very strong incentives for developers who are not bitcoin- or ethereum-wealthy to launch new and competing protocols with their own scarce token.

    1. Michael Elling

      The P2P camp believe they can upend immutable forces. Value always gravitates to the core and top of networks, cost to the bottom and edge. Networks are not flat and bounded. They are infinitely complex and reliant on other networks; much like the human corpus. Signals and incentives are critical for smooth interworking. There are tradeoffs between layers 1-3, 4-6 and 7. There are marginal differences everywhere. Need much more academic work on digital (inter)network theory. It simply doesn’t exist because of our flat and bipolar view of outcomes; purely centralized vs purely distributed.

  11. Cryptoeconomist

    Smart contracts, AI and robots are all invented to serve the interest of humanity at large. Human should not become a slave of their own inventions, especially when the inventions are still in beta.A heist is wrong online or offline.Some critics compare the hard fork to a financial bailout. When the Fed decided to use tax payer’s money to bail out banks in 2008, was there a public, transparent debate on pros and cons? Were the taxpayers given a chance to vote?

  12. pointsnfigures

    Agree. Many crypto currencies and block chains. Arb markets between them to set value. I have followed this and thought if these blockchains are going to be serious competitors to things like credit card clearing mechanisms, exchange clearing mechanisms etc, block size had to go up by a lot to handle the loads.Just remember everyone, free markets are messy. But, they come to more efficient outcomes than autocratic bureaucracies-even if you don’t happen to agree with them.

  13. Jude

    The DAO fiasco tells me that we need a way to fork smart contracts without touching the blockchain. This is not possible on Ethereum, and I doubt it will be possible anytime soon since smart contract deployment and execution are part of its consensus algorithm by design.

    1. William Mougayar

      We need better testing tools for smart contracts, agreed. but they are intricately tied to blockchains.

  14. ShanaC

    I think by far and away the best commentary I’ve seen on this is http://www.forbes.com/sites…If he’s right, this is a bailout as much as code splitting. Why are we not talking about it like that. And he points out that the code should have been more rigorously tested to prevent a bailout in the first place. OyAlso, why are there no economist-cs people. That would be super helpful to understanding what is going on.

  15. Kevin Hill

    I agree with just about everything you said, just not what it means.I think hard forks are inevitable, that adaptability is important, and these recent events probably do signal a bit of strength for Ethereum as an organization.But, I think those stances show a large weakness in blockchains as a technology. It is a tech that takes a lot of costly compromises in energy use and speed to try and ensure that the blockchain is immutable. If you relax those hard immutable constraints, there are lots of inherently more efficient takes on blockchain.

    1. Michael Elling

      “there are lots of inherently more efficient takes on blockchain”. Kevin, can you elaborate?

      1. Kevin Hill

        Sure, I’ll give it a shot.The core problem that the blockchain solves is consensus WITHOUT trust. It achieves this by forcing saying anything to the blockchain, true or false, to be very very computationally and energetically expensive through cryptographic proof-of-work.But, the controversies around Ethereum and hard-forks of Bitcoin show that even with the moment to moment trustless-consensus needed to update the blockchain, there are larger time scale requirements of trust in organizations that can’t be solved by the proof-of-work concept. So, the benefits of trustless organization in the blockchain are drastically undercut by this long-run need for trust, while the costs of such a system are not reduced at all.If you want consensus WITH trust, there are lots of schemes for this. Both centralized and decentralized. On the centralized side, we have something very similar to modern day credit card companies where we enter a controlled network to validate a transaction. On the decentralized side, it probably looks something like an automatic auction to insure the transaction, similar to how online ads are sold, or maybe a reciprocity based information sharing system similar to bittorrent.

        1. Michael Elling

          Thanks! I’m figuring out how best to implement inter-network settlements both north-south (between layers) and east-west (between actors) that provide important price signals serving as incentives and disincentives. These are missing in the “IP stack”. Settlements should result in relatively open, generative and sustainable inter-networks as opposed to today’s siloed world at edge and core.

  16. Ponzi Unit

    Assuming there are hundreds if not thousands of blockchains based on ‘private, consortium and public’ blockchains then if each of those blockchains has events once a year there may be a fork of at least one blockchain, somewhere, on a daily basis.

  17. iggyfanlo

    Super interesting post. I think that this is a really tough one.Ethereum and its fork gives comfort to technorati and those that understand software has bugs and needs changing… so for early, sophisticated, trusting adopters (and perhaps those that lead) this solves problemsBitcoin with its immutability gives comfort to those that feel a financial system needs stability above else, particularly when it’s still gaining credibility.Question of prioritizing trust versus scalability… for software I’d personally (and feels like most of this forum) would choose scalability, but I wonder how the average consumer would feel. For finances, is the average at the point of enough trust that scalability trumps trust?

    1. fredwilson

      That’s why choice and competition is good.

      1. iggyfanlo

        True dat

  18. Dennis Mykytyn

    Of course this decision was wrong. You now have a situation where any action can be reversed by a majority vote. As far as I can tell, nobody is even claiming anything illegal happened, as no laws were broken.When private banks failed before 1933, pre-FDIC, the depositors lost everything. Look to the FDIC as a model, except make it private. Start up a private crypto-currency insurance firm, that insures deposits at Coinbase, for instance, and charges a modest ongoing fee based on deposit size. Get Warren Buffet to be the deep pocket, he loves to insure risk. If you want to start attracting real interest, you have to have insurance to protect depositors from losing everything to theft or fraud or corporate failure.

  19. brian_aaargstrong

    Bitcoin has not EVER had a hardfork. It has had a softfork, and had two competing chains that were much longer than usual, where one simply won out. THOSE ARE NOT HARDFORKS. Stop spreading the completely fabricated and BS narrative that a hardfork has happened before. It is spread to justify a minority fervently pushing for one NOW. It HAS NEVER HAPPENED.How about you actually make your arguments based on facts instead of bullshit you pull out of your ass?

    1. fredwilson

      How about you learn some manners and use your real name

      1. brian_aaargstrong

        How about…I don’t, and I don’t. Because my mannerisms do not change the truth(There has never been a hardfork), and two, its the internet. I can present myself under whatever pseudonym I want, as it is MY RIGHT.EDIT: On a second note, it is beyond pathetic that people like you feel you can dismiss arguments, facts, and points of view on the basis “You don’t use your real name.” What you are essentially saying is “I refuse to deal with reality unless I can combat it by tarnishing your name and reputation.” It is more cowardly and childish than anything anyone can say or do using a pseudonym.

  20. billbarhydt

    I see it slightly different:bitcoin network is a decentralized and programmable asset class.ethereum network is a decentralized application deployment platform.

  21. p-air

    “The Ethereum core developers, after getting a vote of support from the Ethereum community”…you do realize how small of a percentage of the actual Ethereum community voted here, right?

  22. Amar

    This was a very valuable and educational discussion. Lots of goodies here! My favorite summary is http://avc.com/2016/07/hard… ..Props to @wmoug:disqus for taking on the hard questions without rancor or pettiness. Props to @ZekeV:disqus @jameshrh:disqus @lawrencebrass:disqus @SixgillBlog:disqus (and many many others) for asking the hard questions fairly.One thing I would like further clarity on – is there a conflict of interest in owning both stakes in a particular smart contract, as well as ethereum the protocol. Also the question of why is there no transparency in the ownership of the DAO contracts?Maybe I am missing something but it looks like the argument being made is “trust that the hard fork was the right thing because the people who pushed for it are reliable and trustworthy” except none of us have any idea if these people have a conflict on interest or what their DAO exposure was. Wasn’t the whole point of blockchain not to have to trust in a few people in the first place?

  23. Juan Garavaglia

    Fred, if “In my mind, adaptability is more important than immutability.” Blockchain is not for you, Blockchain can’t adapt, immutability is the reason to use any Blockchain, the world is full of databases if you need adaptability is a well developed industry.

  24. Lee Blaylock

    Neophyte on Hard Forks here (interested in learning more about blockchains to I do pay some attention, but have no horse in the race) so pardon the Q if not a smart one, but how is a hard fork different than quantitative easing?From the desc above, they sound highly similar.and if you do or don’t know about QE, then this video is roll on the floor funny. 6mm views. https://www.youtube.com/wat

  25. Alex B.

    In 2008 Satoshi spelled out the rules and game-theoric incentives allowing probabilistic circumvention of the Byzantine Generals Problems. His theory has been proven to work in practice for the last 8 years, a formidable feat I’m sure you will agree.Ethereum set out on a much more ambitious endeavour from a technical standpoint and while already standing on shaky grounds has now decided to change the rules of the games in the beginning of the 1st quarter so as to make them softer to social pressures. This inevitably spells doom for its future ability to withstand these same pressures once they decide to attack, rather than defend, the network.It is a very simple choice Fred:On one hand you have adaptable monies (otherwise known as fiat) which, if history is any indication, typically don’t last more than a few decades, a century at most.On the other hand you have immutable assets the likes of gold that carry with them a 5000 year track record.The question participants of this ecosystem should ask themselves is: do you intend that your blockchain stand the test of time?

  26. brian_aaargstrong

    I let anger get the best of me Fred, but point and case, there has never been a hard fork in Bitcoin: https://t.co/dHhxe8rTnGI realize you are not a coder, and your education is mechanical engineering, as opposed to electrical or network engineering, but the fact remains you are spreading a falsehood as a fact when comparing Bitcoin to another blockchain project. This creates a false impression of the structure, and behavior of, the Bitcoin network in comparison to something else that has taken an action I consider very risky and contrary to the intended function and structure of a blockchain system.Our differences of opinion about Ethereum aside, Bitcoin has not hardforked. I apologize for my angry comments previously, but as someone who works hard(and yes I do troll, that does not mean it is all I do) to ensure factual statements are propagated over misinformation or misconceptions, seeing things that are not true said by someone of your position and reputation can be incredibly infuriating.

  27. bernardlunn

    The hard fork was a tough decision, but the alternatives looked worse. This shows strong governance. Vitalik looks like a good leader.

  28. Gregory Magarshak

    May I just throw my hat in the ring with a different perspective: the problem is that these distributed ledgers are still global. Problems like this could have been much smaller if we didn’t have everyone putting all their eggs in one basket.

  29. Grumpys view

    its the end of digital coins when you can rollback transactions. Period.

  30. Charles

    I feel like I’m missing something. Is Ethereum hard forking because a contract failed to execute as coded, or succeeded in executing as coded?If it failed to execute as coded, I get it. If it succeeded, this seems like a fatal blow to Ethereum’s long-term success and legitimacy. And that seems bad, no?

  31. JamesHRH

    Yes, on a wildly parallel tangent, the big rage here in Canada with our PM the Global Hottie is electoral reform.Oddly, a 15% change in popular vote gave him a 5 year mandate to work his magic.In 5 years, I would like the opportunity for a 15% change in popular vote to remove him from office and give someone else the reins.The three best outcomes for democracies are:- reliable removal of a government that becomes unpopular after having time and a mandate to govern- government by consensus / minority governments in a parliament (not even a close second)- direct decisions by the electorateSome ideas from 300 years ago are still the best ideas.

  32. William Mougayar

    Well, both decisions were hard ones (as V also pointed out), but one had to be taken. Philosophical positions aside, the hard fork doesn’t change anything about the long term characteristics of Ethereum, except that it denied the DAO attacker from keeping what is not his. In my opinion, that is a more important objective than ensuring that the DAO holders weren’t going to lose their money. Decentralization and decentralized decision-making is messy, but it is my hope that we make it less messy going forward, not more messy.