Hacker News Reader

Hacker News

Comment by KirinDave | Hacker News Reader

Full thread

KirinDave

What exactly is the specific use case where Proof of Work blockchain are appropriate?

Consensus in that design is:

1. Perhaps the single most expensive consensus algorithm ever designed.

2. Deliberately one of the slowest.

3. Does not actually provide trustless consensus once 1/3 of the miner nodes decide to collude.

Other than "our data is a merkle tree but you have to trust us to submit transactions to it", and "a currency system with a failed promise to deliver either scale or trustless consensus" what, is the proper application?

View on HN

This reads like a rhetorical question – 'what could possibly be the application for something this deliberately bad?'

I won't speak for what "blockchain" can or will become, but you might see the hype more clearly by considering the ideology inherent in asking specifically for the "proper application." You imply that you or someone in a position you understand might "apply blockchain" to a "problem," as you commonly understand it.

I see "blockchain" more as a movement, which completely evades "application" in the common sense. People building these technologies, and organizations around them, live in a separate world that makes little reference to the broader world of which most of us are a part. They start with distinct priors, which rule out most existing solutions on the grounds that they demand political centralization (see: https://medium.com/@VitalikButerin/the-meaning-of-decentrali...).

You can summarize this ideology as: 'if there's a sysadmin, the system is broken.'

It may very well be the case that they are creating a new world that will be marginalized and outcompeted by incumbent systems and organizations, but that's how I understand what's 'actually going on'.

I admire the principles of decentralization, but I think it's rather like "serverless" - it moves the functionality around in order to pretend. All the cryptocurrencies have pretty "centralized" dev teams; as with Etherum they can fork to decide what history is and most miners will go with that. History is written by whoever has the github.

Consensus forking should always exist as an option. The problem with PoW blockchains is not the visible governance options. Those are a very positive feature.

The problem is the invisible governance patterns that an unreliable consensus algorithm creates.

>History is written by whoever has the github.

Not really, just because a change is pushed to the github repo, doesn't mean the network (and its participants) will automatically accept it. Those unhappy with the changes can always fork the network if they want.

As seen in their very example: ETC exists distinct from ETH

But that never in reality happens. All the forks of Bitcoin or Ethereum are mostly irrelevant. The majority will blindly follow core devs in both cases as have been proven by a long series of forks that became irrelevant very quickly as most people stay with the core version (Bitcoin or Ethereum Core Devs).

that wasn't the case with the segwit rollout on bitcoin. despite the feature being merged and released, miners were able to hold up adoption for months

I think a world where multiple, visible sysadmins compete is much better than a world that pretends to have none when in fact their technology enables invisible, undetectable but exceptionally powerful groups to form out of self interest. Wouldn't you agree?

Are you referring to miners? They are neither undetectable, nor exceptionally powerful. They do have an environmental impact, but the bitcoin community believes the juice is worth the squeeze.

I think the reference was to invisible powers controlling your news, search, social, legislative, and economic institutions.

No. It is possible for miners to collude and get unfair advantages, without even close to 1/3 the hashpower, and without being easily detectable.

I don't know why this is getting downvoted. The current assumption that colluding miners are detectable is absolutely false. There is no way of knowing whether a block comes from a specific mining farm/network as all information about that is forgable by the creator of the block.

If you try to mitigate this by controlling who has access, you just eliminated the advantage of being trustless and thus made blockchain an expensive database.

Please go read the selfish mining paper.

I prefer a world where some Giancarlo Devacini guy, living and working in offshore country, can issue billions of "dollars" out of thin air and being literally unaccountable. /s

Or perhaps they won't be outcompeted. But I think you are right, these are world views in competition.

Or perhaps they are all just blinded by greed and have invested far more than they had to lose... probably much simpler explanation

> 'if there's a sysadmin, the system is broken.'

Am sysadmin, is this best practice or just us being reciprocally lazy again?

> It may very well be the case that they are creating a new world that will be marginalized and outcompeted by incumbent systems and organizations, but that's how I understand what's 'actually going on'.

I keep telling people exactly this! Just because X makes sense or is the best doesn't mean the crowd will agree with it's application. See also all politics ever as an example. I feel like I can say it because I identify as one but, It's just another libertarian fantasy.

I have been asking that for years without getting a good answer. It's a really cool technology! I wanted to use it for something!

But cryptotulips aside, every real-world use case I've looked at can be solved some other way (e.g., real-time published transaction logs). Humans are pretty good at creating trust contexts (see, e.g., 99% of commerce), and they're quite averse to doing anything significant in truly trustless contexts.

Almost nothing is a unique solution to a set of problems. Just because the problems that distributed ledger technologies address can be solved in a different way doesn't mean that to do so is the right business decision.

Sure, but PoW is (by design) computationally expensive, so any blockchain solution has the burden of showing why it's sufficiently better than the alternative to merit the extra cost.

Another problem is that if your nice blockchain has N computers in it, someone can go to AWS and add N+1 and take over your network. You basically have to blow more money on PoW than the network is worth _constantly_.

Or you can have a "private blockchain", which defeats the purpose entirely.

This would be a much better defense if PoW blockchains solved ANY problems.

As far as I can tell, they provide inferior guarantees AND performance to existing methodologies and algorithms.

Is there some other technology out there creating an immutable supply token better than Bitcoin that I don't know about?

What problem does that solve? Fiat money is clearly better than Bitcoin, easier to spend and use.

Yup. The only fields where Bitcoin et al have an obvious advantages is scams, fraud, speculation, money laundering, and a little light drug crime and extortion. Bitcoin has many theoretical use cases, but possibly none that are legitimate, socially useful, and in practice better than alternatives. And we've had nearly 10 years for those uses to appear.

Also: it isn't particularly efficient at these tasks. Proof of work mining has been discovered to have a major flaw AND it doesn't scale well.

I'm sorry, I don't know of one. But that doesn't mean PoW blockchains are actually reliably doing this either.

Any database that only allows appending, and is public readable?

Digital signatures have been a thing for ages.

Sure. And since this is an obvious fact, I figured most people would read that in. But just to be clear, I mean "solved more effectively some other way".

Blockchains in particular carry a business penalty for a few reasons. One is that they're novel. As Dan McKinley says, we should "choose boring technology" [1] unless we have a strong incentive not to. Blockchain expertise, to the extent that it exists, is bid up in the marketplace, giving another excellent reason to pick something else. A third is that the vigorous blockchain hype has created a fair bit of blockchain skepticism in reaction. I know people looking for work that won't even talk to blockchain companies. And then we have questions about security, efficiency, reliability, and all the other operational concerns that one mainly learns through hard lessons (hopefully other people's hard lessons written up for the rest of us to read).

[1] http://mcfunley.com/choose-boring-technology

If you can invent a reliable alternative to proof-of-work distributed consensus there is a multi-billion dollar opportunity there. It’s inefficient, but before bitcoin came along there was no way to even do it. An inefficient way is better than no way at all.

Some other cryptocurrencies are trying alternatives. But none have been as “battle hardened” as bitcoin.

Follow the core developers and you’ll see their focus is on creating an open distributed platform. It’s up to us to figure out what to use it for.

Can I also invent a system that's been proven to not safely guarnatee consensus for years, and trust everyone to ignore it?

If so, I've got a thrilling investment opportunity for you.

> An inefficient way is better than no way at all

Really, why?

IMHO, there's a pretty compelling argument to be made that Bitcoin and the other cryptocurrencies have had a significantly net-negative impact on the world, from facilitating scams and other illegal activity to now having to worry that sites we visit will monetize by stealing CPU cycles to the significant impact on climate change, these technologies have caused a lot more harm than benefit. And all the while, they haven't even figured out how to build a ledger that can scale to day-to-day transactions. My bet is that the whole thing ends up being a tax on people who've forgotten their 17th century Dutch history.

We should've drawn the line in the sand back with 56 bit key limits

> from facilitating scams and other illegal activity

The same was said over and over on the nascent days of the internet, about how it facilitated anonymity so easily and how it would be used by criminals.

> by stealing CPU cycles

Difficulty in bitcoin mining is raising all the time (because of the competition among miners). This is why nobody mines nowadays with CPUs or GPUs but with ASICs. The fact that some sites are trying to achieve some decent mining hashrate by pulling the amount of visitors of a website is a naive move that will soon be off-set by even more raising levels of difficulty. Soon this will not be profitable at all, even if your website was the 1st in Google's page rank.

> to the significant impact on climate change

I like that skeptics raise this point, because:

1) The cryptocurrency mining industry is the fastest one to chase energy-efficient solutions. As soon as a device consumes less electricity, it's adopted widely because of the huge incentives to do so (e.g. CPU->GPU->ASIC).

2) Compare the electricity spent by the entire banking industry with the electricity spent by the entire crypto-mining industry. Now guess what's more efficient.

> The same was said over and over on the nascent days of the internet, about how it facilitated anonymity so easily and how it would be used by criminals.

There's 2 things I want to say about this comment:

1. Those accusations were not wrong. In the same sense that we have pretty clear evidence that a lot of the value injected into bitcoin and stored there is via money laundering.

2. Unlike proof of work currency blockchains, the internet clearly delivered and had no direct theoretical challenges to its very existence that people subsequently and fervently denied.

Proof of work is a bad idea. It's been discredited. It doesn't do what people thought it did. And it's MUCH too expensive to keep doing.

Proof of Work is the only way for people to transact in which they don't have counterpary-risk problems before transacting.

Before PoW was invented, there was no way of doing this.

Now, thanks to layer2 technologies, PoW will start becoming much more (orders of magnitude more) energy efficient.

How is that a bad idea? Make something possible, and later make it sustainable. So long as there's demand for that something, then it's a good idea.

There are two sides to a trade, and typically Bitcoin only handles one of them. Buy some food with Bitcoin and it still might taste bad or make you ill.

Even for purely financial transactions, any kind of loan or investment still has counter-party risk.

So even if a crypocurrency worked perfectly, it doesn't solve very much.

> There are two sides to a trade, and typically Bitcoin only handles one of them. Buy some food with Bitcoin and it still might taste bad or make you ill.

You have this same problem with any form of money. The problem I raised is counterparty-risk when holding your assets (before transacting).

> Even for purely financial transactions, any kind of loan or investment still has counter-party risk.

Yeah, that's why maybe loans should have higher interest (right now, the low interest is artificial, because the bank is really making more money from the fact that it's working as a fractional reserve, not from the interest it gathers from loaning you money).

> So even if a crypocurrency worked perfectly, it doesn't solve very much.

It solves just one thing. Granted, it's not a silver bullet. But there are never silver bullets. Every single invention that humanity achieves, just improves the world a little tiny bit. Cryptocurrency does this as well.

It does not solve anything. It's true that cash and bitcoin both avoid a narrow slice of counterparty risk. But that's not a whole problem, just part of one. As a currency, Bitcoin is an abject failure. Even big proponents admit that: http://avc.com/2017/08/store-of-value-vs-payment-system/

Bitcoin made no practical dent in existing legal financial transactions because it's not better for the great majority of use cases.

Because PoW doesn't have an answer to selfish mining attacks?

> Proof of Work is the only way for people to transact in which they don't have counterpary-risk problems before transacting.

> Before PoW was invented, there was no way of doing this.

If I pay with my credit card and am scammed, I just let them roll back the transaction and let the other side prove their case.

With Bitcoin, that mechanism is _impossible_.

We're not talking about Bitcoin here. I've personally created cryptoassets on the ethereum blockchain that support enforcement freezing and reversing transactions.

The fact that Bitcoin doesn't is primarily down to the ideology of early adopters, not technology reasons.

Genuinely curious: how much does it cost per transaction and how many transactions per second can you execute?

It always puzzles me when people point out that cryptocurrencies are used for illegal activities.

Cyptocurrencies were not designed to facilitate illegal activities. Yet every fuddite makes it seem so.

Every single fiat currency in the world is/can be used for illegal activities. Do we discard them? Care to wager which currency, USD or BTC; is used more for illegal activities? No, I didn't think you would. But media and fudders will repeatedly point out that it's cryptos that are used to transact illegal things.

> The same was said over and over on the nascent days of the internet

Do you shysters read from a script or something? Is this the counter argument you guys always use to try to squelch criticism?

Bitcoin is more than 10 years old. That is ancient in tech. It has yet to be used for anything besides an excellent vehicle for running scams.

>Bitcoin is more than 10 years old. That is ancient in tech

It's also a blink of an eye for a paradigm shift.

> If you can invent a reliable alternative to proof-of-work distributed consensus there is a multi-billion dollar opportunity there.

But Bitcoin only does 7 transactions per second. You can't say it's an alternative to something extremely larger.

The efficiency of mining on CPUs doesn't matter if you don't pay for the electricity.

> My bet is that the whole thing ends up being a tax on people who've forgotten their 17th century Dutch history.

Tulip Mania lasted 6 months

I have a sneaking suspicion that if tulips allowed masses of investors in less-than-open markets around the world to stash/protect assets from their overbearing governments, as well as facilitate the purchase of AK47s and money laundering, that the online tulip trade would be ongoing.

Fundamentally: people aren't buying BTC because online coins are pretty, they're investing in an alternative distribution model whose value is 'fixed' by willing parties on the other side who also rely on that distribution model. Yes, it is the Tinkerbell Effect, but a Tinkerbell Effect backed by centuries old markets of surprising strength (weapons and drugs, just to start with).

The value isn't the thing being exchanged, it's the ability to exchange.

Or you can just use graph theory to make PoW efficient & more secure[1][2]. Skip to page 11 of [1] for a quick visual summary.

[1] http://kadena.io/docs/chainweb.pdf

[2] https://www.youtube.com/watch?v=p1OFJOrgRWw

This looks really interesting. Thanks!

> creating an open distributed platform. It’s up to us to figure out what to use it for.

This is putting the cart very much before the horse.

I have no objection to people researching, experimenting, playing. Godspeed. Spending billions creating a technology that sounds cool but may be useless, though, is absurd.

We didn't get the Internet because somebody said, "Hey, let's spend billions wiring up a lot of computers and see if it's good for anything." The Internet's growth was driven by an expanding set of very specific use cases.

> If you can invent a reliable alternative to proof-of-work distributed consensus there is a multi-billion dollar opportunity there.

Trust.

Just use any authentication/authorization solution, contracts, the rule of law, reputation, everything we've always used to deal with varying degrees of trust.

Utter lack of trust is not a problem people have.

If you want to apply blockchain technology to anything real world (as opposed to coins that only exist _in_ the blockchain), then you need trust anyway whenever there is interaction (e.g. when someone inputs data, you need to trust its true anyway).

It's really only useful for coins, that are themselves less useful than existing money.

It was an interesting mathematical puzzle to solve, shouldn't have been made reality.

This is the right answer. It's frustrating that folks can't see it.

And how is that working out?

I agree that it doesn't seem useful outside of money, but when it comes to money, feel free to trust this. I don't: https://global.spdrs.com/content/dam/spdrs/etf-blogs-images/...

Just to restate because I was not expecting to have to have this post pop up to top of thread: The selfish mining paper shows a significantly worse result than 1/3. I feel the need to lay this out very plainly because folks may not be in on that joke.

It's a joke at 1/3rd because that's sort of a naive threshold of for the BGP anyways.

Bitcoin was supposed to provide a stable solution where distributed actors could come to consensus on a ledger. That's how it was sold in the original paper. Majority attacks against the system will always exist, but the paper mentioned proves that the Bitcoin proof of work algorithm is much worse than we realized, and (more troublingly) that bad actors are harder to detect than it was passively assumed.

In my opinion, the most notable innovation of Bitcoin is the incentive and PoW-reward for the continued upkeep and health of the network, and I often wonder why so much of the literature around blockchains misses or glosses over this fact.

The scheme neatly solves the free-rider problem that plagues other distributed systems, where your continued usage of the network can be ensured by your active participation within it.

>The scheme neatly solves the free-rider problem that plagues other permissionless systems.

FTFY

A parallel example would be various P2P filesharing protocols which was plagued by freeloaders from day 1, and had to come up with various schemes to regulate the user's contribution. The simpler solutions are not exactly Byzantine fault tolerant (user profiles in ed2k) whereas the more elaborate mechanism typically requires a centralised record keeper(private trackers, etc).

A lot of the inefficiencies of PoW are necessary to keep the network permissionless, and conversely you don't need PoW if the system has any kind of entry requirement.

I think blockstack is one such application: to register an identity without relying on any central party for that. Basically anything where you need to make a very small number of transactions to store information that will be cryptographically secured in a decentralized manner.

> Does not actually provide trustless consensus once 1/3 of the miner nodes decide to collude.

Source? This does not match my understanding of POW blockchains.

https://news.ycombinator.com/item?id=16898463

I found this article related to the attack in that paper https://bitcoinmagazine.com/articles/why-bitcoin-mining-pool...

While I'm not completely convinced that the countermeasure of "spy mining" completely mitigates the attack, it does appear that this is not a new vulnerability. Further, if a pool had all of the hardware under one roof, it wouldn't be possible to spy at all which would reinstate all of the problems outlined in the paper.

I don't see how "spy mining" actually helps. It seems a bitter pill to suggest that now we need ANOTHER layer of validation on top of the existing protocol, further increasing Bitcoin's energy cost.

And if you believe this mitigates the attack, then "spy selfish miners" must also exist then, keeping parity.

Proof of work as implemented now is fatally flawed, and it's primarily a tenant of faith instead of logic that it isn't. Just google "selfish mining debate" and enjoy the plethora of inane and tedious reactions.

But also: yes the paper is not at all new. After replication and other checks, the Bitcoin community has gotten up on its high horse, plugged its ears tight with wax, and started signing, "I can't hear you!"

Personally I'm skeptical of all blockchain hype but as I have people in the industry close to me I try to keep an open mind and at least understand the technical and game theoretical aspects of the space. I can definitely agree that the aligned interests of those pressing for adoption and increasing prices has made rational debate difficult in many circles.

I mean, you can call it bad, but it provely works for the thing that it intended to solve.

That thing being sending censorship resistant, electronic financial transactions.

If you don't believe me that this worked, then please go back to 2010 and tell me a better way to donate money to Wikileaks.

At the end if the day, I could send money to Wikileaks, and there were no other options that worked at the time.

The points that you brought up don't really matter, because Bitcoin has been around for 8+ years and we don't see those attacks happening.

I don't care about the theoretical. I care about the practical. Come back to me with that attack vector once it actually happens on the live network, and money going to Wikileaks or whoever ACTUALLY gets censored.

> I mean, you can call it bad, but it provely works for the thing that it intended to solve. That thing being sending censorship resistant, electronic financial transactions.

But it doesn't appear to do so.

1. It cannot meet demand, so folks are recreating existing point to point financial networks and then using the chain in the event of reconciliation. These define a new topology as surely as credit card companies redefined the topology of modern digital economies.

2. It appears that it is very easy to gain a secret majority and do all sorts of nasty things, in addition to freezing out miners.

P.S., if I had a magic wand that could make WikiLeaks disappear in a sweeping curtain of sulphuric fire, I'd stick a bell on the top and wave it happily until my doctor informed me my RSI was fatal.

OK, so it appears that you don't support tools that allow people to get around censorship. That's fine, but that is literally the point of Bitcoin. You just don't like the use case.

At the end of the end, it has been working for 8 years, and it is the best tool out there to solve the problem of censorship resistant financial transactions.

This is proven by the fact that even though people want to shut down payments to places like Wikileaks they are still happening So the purpose is to subvert people and groups like you, who want to stop these payments, and yet the payments and money still flows.

The nasty things that you claim COULD happen, AREN'T. This is proven by the fact that groups like Wikileaks still recieve their money.

My definition of "working" has nothing to with theoreticals. It has to do with the practical, historical examples. And the historical example is that these groups are still receiving their money, so by definition it did work.

Or in other words, if it really doesn't work, as you claim, the how come it its working? As defined by the fact that it is literally the best existing tool out there for this usecase. (IE, visa and mastercard will block your money, and yet bitcoin still works)

If you disagree, fine, go ahead and try to stop my money going to places that you don't want it to (ex, such as Wikileaks). Report it to the NSA or whoever, for supporting "russian propaganda", or whatever. Betcha won't succeed.

> K, so it appears that you don't support tools that allow people to get around censorship. That's fine, but that is literally the point of Bitcoin. You just don't like the use case.

You cannot try to shame me into saying a technology that's been proven to have a major security flaw doesn't have that flaw. It's precisely because I care about this more than you that I can accept that bitter pill and you cannot.

> My definition of "working" has nothing to with theoreticals

Then you are a sucker, and you deserve to lose what you put in. Still, I hope you'll be okay, mmkay?

> Or in other words, if it really doesn't work, as you claim, the how come it its working?

Well, aside from the fact that its not working very well at all... It's prototype-level technology escalated to national limelight because of initial interest from a lot of questionable and probably illegal people. Unlike previous things like the internet, the value exchange mechanism is very direct, making it much more attractive for money laundering.

> As defined by the fact that it is literally the best existing tool out there for this usecase. (IE, visa and mastercard will block your money, and yet bitcoin still works)

I'm fairly sure the cost of a dozen viable stolen identities to use as money movement channels is MUCH cheaper than the transaction fees on bitcoin, so it's not like the enterprising criminal doesn't have a classical, high performance, downmarket option.

Bitcoin in particular is libertarian chic and anarchist crypto-religion. Ethereum has MUCH more interesting properties, but still needs several years to bake.

> If you disagree, fine, go ahead and try to stop my money going to places that you don't want it to (ex, such as Wikileaks).

If you really wanna donate to a bunch of authoritarian-loving pro-rapist people with such bad opsec that a single twitter actor blew a core comms network wide open? Be my guest. Crypto-fash it up.

> I'm fairly sure the cost of a dozen viable stolen identities to use as money movement channels is MUCH cheaper than the transaction fees on bitcoin, so it's not like the enterprising criminal doesn't have a classical, high performance, downmarket option.

Ok, and in 2010 did the average person have access to such a network?

I personally didn't have access to those things. Thats sounds pretty difficult, actually. But what I DID have access to was bitcoin. And it worked! And it didn't get censored like visa or mastercard was doing.

Thats the usecase. Me sending money to wikileaks. I cannot think of any other plausible way for me personally to send money to wikileaks back in 2010. Stolen identities sounds like it would get me sent to jail, and it seems hard. No thanks. Bitcoin was easier. (A dollar for a transaction fee is still cheaper than cross country money transfers, actually.)

> Ok, and in 2010 did the average person have access to such a network?

Yes. It was called IRC.

>1. It cannot meet demand, so folks are recreating existing point to point financial networks and then using the chain in the event of reconciliation. These define a new topology as surely as credit card companies redefined the topology of modern digital economies.

Are you talking about the Lightning Network?

While it has a different topology and its own limitations, I wouldn't call it a replication of existing systems. It's an extension of the blockchain transaction concept more than anything.

Users can still make trustless payments without relying on a third party. The routing nodes do not require trust due to the hash time locked contract system.

> Users can still make trustless payments without relying on a third party.

You need to establish each tunnel on chain, as I read it, this requires a commit to the block. So any time you want to open a payment to a 3rd party you need a btc chain transaction. I'm happy to be wrong here, but that's how I read it, and if so then that's not really gonna fix the problem at all; it's just going to make more not-actually-decentralized services.

Maybe it's an okay thing to imagine a world where there is a primary chain for major institutions and then brokers on this chain using lightning. But I gotta confess I was sorta signing up for something more revolutionary than "disrupting visa with a new visa."

Lightning network payments can be routed through other users/nodes. So you only need one "on chain" transaction to then be able to send those funds back and forth to any other user in the network.

eg: Alice opens a channel with Bob. Charlie opens a channel with Bob. Alice can send payments to Charlie by having them routed through Bob (in a trustless way using Hashed Timelock Contracts).

There are valid concerns around what the topology looks like and how efficient it is to find optimal routes through a large network. These are the problems the developers are working on though.

"Just put your life savings on chain and we'll solve the traveling salesman problem?"

Or am I misunderstanding?

Haha. Well that's why the amounts of money people are putting in the network right now are very small. It's certainly not something anyone recommends using for large payments, at least yet.

I think many people envision it something like an account balance you top up as you need to.

None of this should have much of anything to do with investments. I don't "invest" in crypto in the same way that I don't "invest" in the Euro, or even in the US dollar.

I put the vast majority of my money in the stock market. And I have some money in other currencies, centralized services like venmo/paypal/ect, and yes, a bit in crypto. Currencies which I then use to spend on things.

The point of a currency, yes even cryptocurrencies, is to SPEND it, not hold it.

completely agree re your argument that the point of cryptocurrencies is to SPEND it. all the hodling isn't making cryptocurrency adoption easy. but also def a lot of work needed for solving scalability issues and making cryptocurrency spending more end-user friendly.

> then please go back to 2010 and tell me a better way to donate money to Wikileaks

Regular SWIFT wire transfer always worked. The blockade applied to VISA/MasterCard/PayPal/Western Union.

The problem is that entry points in the network (exchanges) are censored or will be and you need exchange to actually extract money to spend, living in a authoritarian country. You can argue that people can use pure bitcoins by themselves, without fiat money, but that means that there should be an infrastructure where organizations accept and process btc, and they aren't doing it due to technical limitations of blockchain. It will only function as a "resistance coin" until there are some unaudited exchanges or some freedom in movement to go to other country for your money etc. If/when country will go "full dictatorship" btc won't be able to provide any useful utility.

> Consensus in [proof of work blockchain] design is:

> 1. Perhaps the single most expensive consensus algorithm ever designed.

It's like arguing that bubblesort is one of the most expensive graph traversal algorithms. It's not, because it's a protocol for document timestamping, not establishing a consensus.

> 2. Deliberately one of the slowest.

Indeed. It was designed so verification of the timestamps is much, much faster than producing them. In previous work, with trusted third party operating a timestamping service, it was enough to control a single service, so there was no need to design this into the very data structure.

> 3. Does not actually provide trustless consensus once 1/3 of the miner nodes decide to collude.

Indeed it does not, because it was never designed as a consensus protocol.

> [...] what, is the proper application [of blockchain]?

When you need to establish an order in a sequence of related documents, like with trade contracts. Outside of this very narrow purpose, blockchain is pretty much useless.

> Indeed it does not, because it was never designed as a consensus protocol.

What exactly do you think a distributed ledger is other than a normal ledger and a trustless consensus protocol?

> What exactly do you think a distributed ledger is other than a normal ledger and a trustless consensus protocol?

There is no such thing among cryptographic protocols as "distributed ledger". The ledger is implemented on top of document timestamping and while required for Bitcoin, it doesn't exist for blockchain data structure. It's like calling HTTPs a cryptographic protocol for serving files.

About this "trustless consensus" part, how do you define "consensus" and how does it relate to what computer science calls "consensus" with its most prominent problem of Byzantine generals, especially with the impossibility proof (the one that talks about 1/3 traitors)?

I'm sorry, but I cannot even work out what your post is about.

Firstly: the single most important function of the Bitcoin network is to maintain a ledger that prevents double spends. Ethereum's charter is not so easily summed up bit it shares a responsibility here.

The tools Bitcoin uses to do this do indeed stem from cryptographic research, but the primary function of a blockchain is not to certify the past but to come to consensus on the immedate future.

No elliptic curve will do that for you. It is a distributed algorithm leveraging guarantees given by cryptographic primitives. Every expert accepts it as such.

As for your final note: multiple "solutions" for the Byzantine generals exist and have been published for years. Many of them give better guarantees than what a post-selfish-mining PoW blockchain gives you.

If your argument was, "It was never possible to come to consensus to begin with" then what are we talking about. If your argument is that it works fine, you should refute the cited paper directly. If your argument is that it's fine despite the flaws, I do not agree.

> I'm sorry, but I cannot even work out what your post is about.

The point boils down to this: you throw the "consensus" term left and right. The term has a concrete meaning in computer science, and you use it incorrectly, so stop doing that.

> As for your final note: multiple "solutions" for the Byzantine generals exist and have been published for years. Many of them give better guarantees than what a post-selfish-mining PoW blockchain gives you.

Note that the Byzantine generals problem has a rigorous proof that says you cannot have a consensus if loyal nodes don't outnumber the traitors by more than 2:1 (so you need 2/3 + 1 loyal nodes), and the proof is not about just the original algorithm Lamport et al. designed, it's about the consensus problem itself.

Where does the blockchain's "50% + 1" stands against that if blockchain was supposed to be a consensus algorithm?

> If your argument was, [...] If your argument is [...] If your argument is [...]

My whole point is "don't use the 'consensus' term incorrectly" and "call the blockchain what it is, i.e. timestamping protocol".

> My whole point is "don't use the 'consensus' term incorrectly" and "call the blockchain what it is, i.e. timestamping protocol".

The time is 35. If you don't agree, we'll have to come sort of... Oh wait I'm not supposed to use that word.

Please provide sources and/or elaboration for/on the following:

* Deliberate slowness.

* Comparative slowness (given the problem space).

* 1/3 collusion breaking trust-less consensus.

* A definition of "us", "trust" and "scale" for your 6th line.

> Deliberate slowness.

As OP article alludes, bitcoin difficulty is tuned for a very long interval between blocks, to reduce the risk of nearly-simultaneous "uncle" blocks.

Other blockchains such as Ethereum have seemingly successful alternative approaches. IMO the article has a very limited perspective on blockchain technology. It's understandable, since the space is exploding and very hard to keep a handle on, but I think it should be less sweeping in its claims.

That doesn't really solve the problem at all though.

Ethereum beautifully solved the problem of "how do I cheaply make a token with trustless trading semantics?", and it's been used to raise billions of dollars of funding. So that might depend on the problem you're talking about.

There's an enormous amount of interesting blockchain technology, so sweeping dismissals of it tend to be lazy and ignorant.

> to raise billions of dollars of funding

How much of that was neither exit scammed nor lost to contract bugs?

Not the point. It's obviously introduced new problems, but does provide a highly valued capability. Sweeping dismissals of the space based on bitcoin's design are ignorant and lazy.

How do we know there isn't a selfish mining ring owning and subtly biasing Ethereum?

I want a good answer to this question before I put money back in. Because tbqf, it's very easy to raise capital on buzzwords. Especially if you're so we'll suited to money laundering.

It should be possible to check for autocorrelation of block arrival times, and correlation with forks. I don't know whether anyone's gone looking for that, but Eyal and Sirer have a strong incentive to monitor for it, and obviously have the chops to do so.

While there are bad actors in every system, how that system handles bad actors is still relevant and to the point when judging it.

Hardly lazy. The entire space has existed for 10 years and the underlying tech for even more. So far all that has been produced is massive amounts of fraud. If you want to count pump and dump scams, Ponzi schemes, pyramid schemes, terrorist funding, organized crime money laundering schemes, crypto-locker scams, “underage kids buying hardcore drugs” schemes, child trafficking schemes, blatant tax fraud, “hiding money from the ex wife” schemes, and all the other underground frauds, scams and criminal behavior as “valuable”, be my guest.

... oh and all at the low low cost of the several small nations worth of electricity required to power the brutally slow, inefficient by design Rube Goldberg contraption that is the blockchain...

The whole space is a joke.

how much of the internet was dedicated to porn and funny pictures?

Over 99% of it?

The PoW blockchain is slow to propagate updates but design. If you don't know how proof of work blockchains work, I advice you to Google around for a nice technical breakdown. Suffice to say, it is fundamental and a property people specifically code in to make blocks take a specific amount of time to complete.

As for the 1/3 citation, I guess that's slightly less known: "Selfish Mining". Citation: https://arxiv.org/pdf/1311.0243

Of course "1/3" is a joke. It's actually closer to 1/4.

TL;DR for those of us on small viewports?

People can collude to get a disrporotionate amount of the profit from a blockchain. The more people collude, the more profit they get, the more rational it is to join them.

This then bridges into traditional 51% attacks.

> The more people collude, the more profit they get, the more rational it is to join them.

This is not true. Colluders diminish the value of their PoW reward by reducing the trust of all other participants in the system and honest actors have an even strong incentive not to join them. Just a few years ago a mining pool that gained a very large share of the Bitcoin hashrate had members choose to leave -- https://www.coindesk.com/bitcoin-miners-ditch-ghash-io-pool-...

Please see the referenced paper. They address this.

Because Selfish Mining attacks are not inherently detectable, you cannot expect the average user to detect them.

Too many refutations of selfish mining appeal to an oracle. This is an example of such an argument.

If the big miners need to self regulate like that, then why bother with proof of work in the first place?

You say, "traditional," but has one ever actually taken place?

Bitcoins are mined by a uniformly random process, which means even groups with a small amount of power can mine blocks, and even can occasionally mine more than one in a row. But instead of publishing that mined block immediately, there is a better strategy. Waiting on publishing makes a fork, and the colluders can keep mining on the fork until the strategy defines that they should publish - which is basically when the "honest" branch catches up to the colluding branch. Since the colluders don't have a majority, this will always happen, but that's ok.

Due to the way that the Bitcoin protocol works, honest miners will sometimes "help out" the colluding group by mining on that published fork. This basically amplifies the power of the colluding group, which is the goal. Worse, the paper proposes that the percentage of honest miners switching to the fork can be influenced cheaply, by choosing which blocks to forward along in the p2p network.

Interesting. Wouldn't this be riskier than non-collusion, since it diverts hashpower from the main chain and the merging of the side chain depends on help from non-colluding miners?

Great paper!

Reply on news.ycombinator.com