Hacker News Reader

Comment by m-i-l | Hacker News Reader

I used to think that the blockchain bubble was like the dotcom bubble, in that the core technology was actually genuinely transformative, so that even after the bubble burst and all the over-valued companies went bust, the technology would remain and continue change our lives. However, I must admit I'm starting to have my doubts now.

It was 10 years in to the dotcom era that the dotcom bubble burst, and by that time we'd had massive maturation of the technology and the start of mass adoption.

However, we're approaching 10 years since the start of the blockchain era, and there are signs that the bubble is bursting already, but in terms of progress it is nowhere near where the dotcom era had got to by this stage, and there's not much indication of anything that appears that it will be particularly long lasting or widely adopted. Sure there's niches like settlement systems and security tokens, but not something that a member of the public is going to use on a daily basis like they do with many of the internet based companies.

View on HN

Maybe we should consider that the blockchain's killer app is the truth.

This is the case with bitcoin, since the purpose of the blockchain is to be, for example, trustless, protecting you from someone whom you don't trust lying about your bitcoin transactions. Bitcoin will show you the truth about those transactions. The blockchain also tries to protect against fraudulent double spends and do on, and this like all fraud is also lying, untruth.

Likewise, the economist Hernando de Soto Polar's work on property rights and "dead capital" has led to his creation of a blockchain app to allow people in unstable political environments to prove their property claims. If the old data is destroyed in the flames of the next revolution or hacked by your neighbor bribing an official, it doesn't matter as much because you have your claims on a global registry, and that global registry contains the truth.

And maybe when machine learning can allow bad actors to create false media, change what a politician has said in a video, perfectly alter or even create fictional recordings of conversations in order to deceive, the blockchain may provide some shelter for the truth.

Edit: Of course there is a caveat: bad data in, bad data out.

That's naively idealistic. In unstable political environments when someone with power and weapons wants your property he'll just take it. No one cares about "truth". And they can use threats of violence or imprisonment to force people to update any property registry to their liking.

Of course de Soto Polar's vision isn't all-powerful but it is worth looking into just the same.

Here's his website: https://www.desoto.com/

And from one of many many articles:

"In the late 1980s and early 90s, de Soto played a key role in ending the Peruvian Terrorist group Shining Path’s violence by getting the Peruvian government to recognize poor property owners’ land deeds."

https://medium.com/transpartisan-notes/blockchain-hernando-d...

Your comment is exactly why I believe many blockchain (and even cryptocurrency) technologies are useless.

There seems to be this underlying pull for many crypto/blockchain enthusiasts that this technology frees us from the constraints of governments and society. "The truth is in the blockchain" and "There is no government in charge" are common refrains.

The thing is, unless you want to be a true hermit, you can't escape the bounds of society. Nobody really cares about the "truth", they care what society says is acceptable (and what governments will enforce). I'm pretty sure there were a lot of contracts between Native American tribes and American settlers that were basically ignored, and everyone knows this, but there aren't any title companies that are worried about this fact.

You're taking an absolutist view. Crypto currencies weren't intended to free people from every constraint of society, just a few constraints, and only to a degree.

Crypto clearly has the potential to make enforcing trade and currency restrictions far more difficult. When it's too difficult to enforce a rule, governments can (1) stop enforcing it; or (2) ratchet up the penalty when someone does get caught as a deterrent. The second option can lead to unpopular disproportionate punishment, and slowly melts away towards the first one (e.g., war on drugs).

Crypto-currency won't take down governments, especially overnight, but they clearly have a long-term potential to alter some levers of power.

I don't wish for myself or others to escape the bounds of society and government (ok, maybe some of the bounds but that's another talk). Indeed my premise is that individuals, society and government must all be bound to each other in order to have a desirable society, and the sin qua non of this is respect for the truth.

As for government, I happen to prefer democracies, and they are especially beholden to truth. When saying that the blockchain's killer app might be the truth, I'm thinking about how it could help save our democracies.

As for The Truth, as in the ultimate truth, for when you'd use quotation marks -- the "truth" -- I'm not going there. We need only concern ourselves with the notion of truth as opposed to lies.

To clarify the distinction, see how this doesn't work:

"Is it true that you were in Symphony Hall at the time of the murder?"

"Is it 'true'?... What do you mean by that slippery word, 'true'? This notion of 'truth' is simply whatever we agree on and therefore irrelevant."

Edit:

And if the truth is simply what is agreed upon, then are lies simply what is not agreed upon?

If so, perhaps we can agree together that the year is 1984.

Nothing of what you are saying is something blockchain technology addresses.

Hi, excuse the delay.

I was saying that blockchain might help address fraud, against untruth or deceit. For example I believe that yes, the American Indians would have been better off with unalterable titles to their lands. Would it have changed history? I don't know, but in other situations it might, and for that I point you to Hernando de Soto Polar's site. https://desoto.com

I was also saying that the claim that "nobody cares about the 'truth'" (and so blockchain is not useful) is misplaced, and that in this case there's no good point for using the word truth in quotes as if it were something illusory. There are times when quotes around the word truth make sense, like in philosophy, but this is not one of them. An exception is if you are talking about the potential to use the blockchain to make lies permanent.

And third I was responding to your statement that people who support the blockchain are often hoping to be freed "from the constraints of governments and society", and that my comment was a good example of this. Your comment might be correct about others, but I'm not a libertarian or anarchist. Rather, I'm a bit worried that technology like machine learning and even just social media have the potential to weaken and even destroy our democracies by using lies (doctored video, for example) as political tools.

But you don't even need machine learning and social media. In Russia, lies have undermined what began as democracy under Yeltsin to such a degree that it is now for all intents and purposes a dictatorship. In Russia the playbook is basically as follows: an event occurs that makes the Kremlin look bad. The Kremlin uses its media outlets to make up alternative truths (lies), several versions, some more implausible than others. The person seeking the truth, what actually happened, is now looking into a hall of mirrors. Not being able to see anything clearly and having to spend their energy disproving several false versions, no meaningful discourse happens. The event, in this way, has been neutralized by lies. If a blockchain can help with this kind of problem, then I'm interested.

Edit: Btw, I'm not like a cryptocurrency enthusiast, don't program Ethereum contracts, own no BTC or any other crypto, and believe and have argued here at HN that BTC is a bad currency because of volatility with makes credit and debt impossible. Additionally BTC is used for crime and as a way to cash in on fraud, malware, threats and so on, and I'm not interested in any of that.

So specifically on addressing fraud, this only helps the merchant. You could make an argument that decreased fraud decreases costs but human greed overtakes that assumption. So then what’s left from the consumer side? What if the merchant send you a bad product? They have no incentive to make it right. You could argue that eventually it will put them out of business from bad reviews however this doesn’t always work in practice and you’re still out of your money for the original bad product. With a centralized system we can just reach in their account and take it, refunding you your money and costing the merchant more in the long run, further making them want to do legit business. This permanent contract you speak of is a double edged sword that cuts the consumer more than the merchant.

If so then the ultimate application of blockchain is authentication of recorded video as actual, unadulterated video of an event.

Why stop at video? Every application that creates any file should take a hash of it and upload it to a public blockchain. You could use it to identify tampering with any individual file. If deployed more broadly (e.g., on the file system level), you could easily identify if your systems have been hacked.

Yes, that is a possible application of "blockchain technology", given a broad-enough definition of what constitutes "blockchain technology".

But the consensus-based proof-of-work (or proof of storage, or proof of memory, etc.) models aren't really necessary or even clearly desirable.

There are timestamping systems dating from long before "blockchain" (at least 1990, probably before) which involve sequential hashes, with each subsequent hash also including the previous one (in some respects this is a very simple blockchain), and then if desired you can roll them up through a directed graph to one master hash produced periodically and published immutably -- printed in a newspaper was one approach, or posted to Usenet, but you could think of others -- such that it validates all previous hashes as having been created before a particular time.

Guardtime was doing this back in 2007 or 2008 and printing the top-level daily or weekly hashes in the newspaper; I noticed they are now marketing themselves as a "blockchain" company, which is astute marketing (and I'd do the same if I were them, probably), but the ability to do secure timestamping existed in several forms without blockchain.

The reasons these systems weren't used have less to do with technical feasibility than lack of pressure to implement them.

> sequential hashes, with each subsequent hash also including the previous one

Isn't this "in principal" what a blockchain is? Just because they didn't call it a blockchain back then doesn't mean it isn't.

You could also consider Git to be a blockchain of sorts.

What distinguishes Bitcoin in particular is proof of work which provides a means by which distributed untrusted actors can transact in a trusted way. The actual chain of hashes just records these transactions.

The answer to that is tricky because it depends on intent.

Taking an expansive definition of blockchain as a log of sequential, cryptographically verifiable transactions seems okay enough. At the point where we're nitpicking what is/isn't blockchain, that's not a useful exercise, because the real question is whether this is the right solution for the right problem.

On the other hand, "blockchain" is a highly loaded term, and choosing to label something on the margins "blockchain" means diffusing the legitimacy of application onto the marketing term, or vice versa. In that sense, I'd be hesitant to lump this marginal example into "blockchain" specifically instead of just "cryptography stuff".

No, blockchain does not enforce anything about the original video. It only enshrines the version published initially by putting it (or a hash to save space) on a permanent public record.

> Maybe we should consider that the blockchain's killer app is the truth.

That isn't any objective truth though, it's merely the consensus as currently defined by the majority of participating processing power.

> Maybe we should consider that the blockchain's killer app is the truth.

That isn't any objective truth though, it's merely the consensus as defined by the majority of current participants weighed by the processing power they command.

It may seem like a subtle distinction, but I think it helps make the benefits and risks clearer.

First: the dotcom bubble burst five years after it began, and the technological advancements didn't see "mass adoption" - because they were not consumer facing. I get the feeling that you're thinking that pets.com was the dotcom bubble... the bubble was centered around telecom technology - not retail websites. The bubble is fairly disconnected from what one would describe as the success of the internet today - web 2.0, the rise of APIs.

Second: The region of experience that blockchain tech is best suited to revolutionize, finance, is also the one that is most sensitive to rapid change and most heavily regulated. The US government is still grappling with the idea, hell - the NIST draft that was supposed to provide some clarity of what cryptocurrencies actually are got totally nuked by every interested party. I'm still waiting for the IRS to issue a clarifying statement on the tax handling of a hardfork so that I can amend last year's filing. It has led to a very strange situation where a hands-off approach has allowed for development, but also prevented integration.

On your first point: I seem to remember plenty of companies that failed that were not telecom. In fact when I went looking just now I found a bunch of listicles about retail startups that failed.

The telecom bubble and crash was almost a side-effect. The telco executives saw the web taking off and used it to justify wildly over-inflated projections of traffic demand growth, and then began building out debt-financed networks to meet it. When the traffic didn't materialize (partly because the dotcom bubble burst), they had a bunch of overcapacity not generating revenue to service the debt that financed it, and that killed them.

> In fact when I went looking just now I found a bunch of listicles about retail startups that failed.

Their contribution to losses in the dotcom bubble were measured in the millions. Telecom, last mile service providers and B2B losses were measured in the billions. [0] You obviously know those lists are infotainment.

> The telecom bubble and crash was almost a side-effect.

I'd characterize it as more of a feedback loop. They weren't simply increasing capacity with the expectation of a customer demand that didn't materialize, they were pushing through an evolutionary hump in search of a global minimum. The world would look very different today if they hadn't, at great cost, setup the network the way it is today - pushing way beyond the original design of NAPs (Network Access Points) to IXPs (Internet Exchange Points). I guess you could attribute the change to capacity, but I'd say it is more accurate to describe it as a recognition of inevitable scalability problems and as a result - a change in philosophy. That change, the official government handoff, occurred in '95. There was also an amazing amount of very interesting R&D work going on, from strange new operating systems to packet switching on fiber optics without copper interruption. Even Enron was getting in the game by trying to setup a commodity market for bandwidth. The entire situation reminds me of the evolutionary problem, where you've got massive metabolic costs in brain size growth.

So no, it wasn't an overestimation of traffic growth. Everybody was trying to figure out the way the new world would work, a lot of people got it wrong.

Also... 401k participation spiked in '95 - I don't think that is a coincidence.

[0] https://money.cnn.com/2000/11/09/technology/overview/

I'm confident that there is significant real value to be gained from blockchain technology but at least 99% of projects today are just empty buzzwords. Most wealthy people in the blockchain space are well-meaning but they don't know what they're doing; they invest their money all over the place based on superficial whitepapers and buzzwords.

I think that it will take several more years before we start seeing useful blockchain applications.

It's possible that the bubble may have to burst before the industry can progress though. Right now, smart people might be intrigued by blockchain technology but most of them don't want to get into the space because of the level of speculation.

Right now, it's just more profitable to launch some generic buzzword coin that does nothing than to launch one that actually does something useful.

Just like in AI right now, there are a lot of people who just want to make a lot of money and they don't care how.

we are -80% from ATH, I don't understand why peoppe talk about the bubble will burst in the future. It happened already.

Bitcoin down 80%?

when it was at 5800 and ATH was around 21000? kind of yes

TFW mods downvote facts

A lot of the utility from blockchain doesn't come from being able to do "new" things. They come from being able to do old things better: cheaper, more securely, free from the interference of trusted parties / government.

It remains to be seen which processes the blockchain will take over. Some are obvious, ie. prediction markets, sports gambling, money remittance.

Some are more complex, and will require frameworks to work within trusted parties: ie. tokenizing physical assets or securities that exist outside the blockchain world. Identity management.

Maybe not government, maybe just humans in general. Decentralisation, blockchain, smart contracts, cryptocurrency, AI and IoT could really start coming into their own further into the future when we start having more autonomous agents performing those little tasks on our behalf, like a smart fridge ordering more milk and interacting with a drone to ensure payment and delivery without having to bother a human.

But I'm not getting the sense that any of the huge amounts of money being invested is going into much other than feeding the crypto ecosystem at the moment.

True story: VCs specialising in funding blockchain projects told a dev team "You guys are building a proper business, that's great but it means it won't 100x in a short period, so isn't really what we're after".

You sound incredulous about this – VCs have a pretty clear business model, funding slow-growing steady businesses is not it.

To be fair, that business model includes obscuring that aspect of the business model as much as possible.

But VCs normally try to fund genuine businesses - in this case, by suggesting they weren't interested in a "proper business" (as in a legitimate business) and were wanting something to exit in the "short term", the insinuation was that they just wanted a sham / fake business that they could use to fool people just long enough to do a 100x pump and dump.

> True story: VCs specialising in funding blockchain projects told a dev team "You guys are building a proper business, that's great but it means it won't 100x in a short period, so isn't really what we're after".

Sounds like that VC firm knows what VC is, and would correctly give that answer regardless of specialization. Nothing to do with blockchain.

>"They come from being able to do old things better: cheaper, more securely, free from the interference of trusted parties / government."

Name one thing that blockchain has made, or is even close to making cheaper or more secure or trustless?

The largest application of the technology has been bitcoin thus far, and bitcoin is more expensive; less secure than the traditional financial system (where if you get hacked you at least have some recourse); and you still have to trust intermediaries like Coinbase to acquire the Bitcoin and Bitpay to actually buy anything with it.

And I know the canned response to that is usually "why can't X new cryptocurrency solve these problems eventually? The internet had its problems at first as well."

But you solve those problems by making bitcoin exactly like our current financial system, with intermediaries and trust and regulation. And minus the hilarious waste of time, money, and energy that is proof-of-work.

I'll give you one.

The year is 2010. And I am a normal citizen and I want to donate money to WikiLeaks without the banks colluding to stop my financial transaction, even though nobody has been convicted of a crime.

How do you do it? The only way I can think of to do this easily (easily being the key word!) is a cryptocurrency transaction.

Banks and credit card companies collude all the time in order to censor financial transactions between people who have committed no crimes.

Do you have a better way of easily sending censorship resistant financial transactions? Because Bitcoin worked just fine back in 2010 for this usecase, even though the banks tried to censor it on their platform.

Most of the people downvoting you probably never had to think about making a censorship free transaction

With bitcoin, what would stop governments from passing laws making certain addresses tainted resulting in people that send coins to or receive coins from them to be audited and having to explain why and how they linked to certain causes.

What makes you think if bitcoins (or any *coin) becomes the main currency in a country or region, we would not have laws were all addresses would need to be linked to people (or companies) and the ledger would not be used for figuring out and persecuting people supporting certain causes?

Well, for one, the government would actually have to make such a law, and then suffer the political consequences.

I would much prefer that the government is forced to actually follow the legal process, instead of what happened in 2010 with WikiLeaks, where the government made vague illegal threats against the banks, in order to censor transactions between people who broke zero laws.

It forces everything out in the open, instead of giving a couple banks the power to made hidden agreements between each other in order to collude and censor perfectly legal financial transactions.

And so far, the government has NOT made such a law. So guess what, that means it works!

This is not about people breaking the law. This is about people following the law, and yet private parties are STILL able to censor transactions through collusion and because of a few vague, and probably illegal, threats from the government.

The fact that the government has so far NOT been able to pass such laws, is innovation in and of itself. It means that it works right now for the purpose of making censorship resistant transactions, because before people WERE succeeding in censoring them, and yet now they aren't!

Read up on money services business and the various regulations they must follow, including the anti money laundering regulations. You may find yourself surprised at what laws exist.

The government can do whatever it wants, the point is that it’s decentralized design makes it basically impossible to censor.

Given the size of Bitcoin's keyspace, what would you surmise is the cost to a particular cause of "tainting" one of its public keys?

Coinbase seems to already do this to a degree. If you transfer to a flagged address your account is banned with zero explanation or recourse. This will likely be the norm going forward for any reputable exchange.

> I am a normal citizen and I want to donate money to WikiLeaks without the banks colluding to stop my financial transaction, even though nobody has been convicted of a crime.

Time will tell if Wikileaks was really a noble cause, or if the banks were right all along.

It is now 8 years later, and yet the organization has not been convicted of any crimes. Time already did tell who was right and who was wrong.

But even IF they are convicted of a crime sometime in the future, this is 8 years later.

Censoring financial transactions should require a court order. Signed by a judge, and done through the normal criminal justice system. And it should be done publicly, so that judges and politicians are held accountable for their actions.

In America, due process is a constitutional right, and if someone is guilty of a crime, there is a process that we should be going through.

Punishment should not be inflicted upon others outside the court system via secret agreements and collusion between oligopolies.

We have laws and human rights for a reason. Don't be so quick to throw them away.

Let's go to slide #2 then-- Scihub. They take Bitcoin which they ostensibly use to pay for the servers that serve up the science journals.

Do we need time to tell if Scihub is a noble cause?

> I want to donate money to WikiLeaks

Trustworthiness of Wikileaks aside, do you think that's a $500 billion market? Donating to activist organizations?

https://qz.com/775159/theres-a-500-billion-remittance-market...

"Worldwide, 230 million people send $500 billion in remittances each year, primarily using firms like Western Union, Moneygram, and RIA, which together control 1.1 million retail locations and account for more than 25% of the world’s annual remittance volume."

That's not an answer to the question above...

It's funny that the downvoters haven't really come up with a valid argument why Bitcoin is not a valid use case for this.

It's much more difficult to prove that person A transferred money to B via Bitcoin than it is via regular financial transactions due to Bitcoin's pseudonymity. Similar it's equally difficult to prevent this from happening.

Just that the you do not seem value in this (censorship-free transactions), does not make the argument invalid.

Not as big as the bitcoin market but corda's growing[0] use for forwards contracts between parties[1].

Though it seems like this is an instance of blockchain "like our current financial system, with intermediaries and trust and regulation", but with cost/time savings by reducing the amount of humans involved in the transaction reconciliation process.

Many more examples of old finance firms using blockchain [2], though a far cry from the whatever is going on in the cryptos in use with consumer discretionary trading.

[0] https://github.com/corda/corda

[1] https://www.thetradenews.com/commerzbank-replicates-fx-trade...

[2] https://www.globalcustodian.com/news/technology/

PKI is now quite trivial. It's easy for anyone to keep a log mapping a public key to some type of identifier.

The public didn't buy most of their stuff online 10 years into the web either, but now Amazon dominates.

Blockchain needs a dominant company to stay relevant. As with Amazon, the entire company doesn't have to be tied to it, it can have onchain/offchain components.

I think this becomes easier as tooling becomes easier as well for developers and for users.

"I think this becomes easier as tooling becomes easier as well for developers and for users." Agree, this is what we're working on.

Banks and financial institutions are also big consumers of blockchain tech. Jibrel is putting fiat, realestate, and other assets on chain, for example, which is convenient for ownership tracking, but banks can also make on-chain/off-chain arbitrage. So while the average consumer can gain benefits from some of these blockchain projects, adoption by big banks, governments, and institutions are a greater sign of things to come IMO.

> we're approaching 10 years since the start of the blockchain era

By that standard, the start of the internet era was in the early 80s, almost 20 years before the dot com crash.

Don't forget that Ethereum has only been live for 3 years. This is still very early technology.

"there are signs that the bubble is bursting already"

The bubble is barely even starting. There might even be another bubble and crash before we get to the real one. These are just practice bubbles.

> Sure there's niches like settlement systems and [...] but not something that a member of the public is going to use on a daily basis ...

I'd argue that settlement systems aren't a niche system and is something most of us use on a daily basis -- if not more frequently. Unless you're using cash for everything, every transaction you make goes through a settlement process, sometimes multiple times.

People often talk about how inefficient blockchain is, but they rarely have a grasp on how horribly inefficient our current settlement systems are. They are slow, expensive, inefficient, and complex.

I have doubts if it'll be widely used outside of the financial space, and I recognize that may be the context in which you were using the adjective "niche", but I do think it has the potential to be transformative within that space, and I think that space is potentially bigger than many of us realize.

One of the points is that settlements systems are private permissioned blockchains, which don't take advantage of one of the main innovations of the public blockchains, i.e. decentralised trust. Interestingly I noticed that the recent Bank of England Real Time Gross Settlement Proof of Concept report[0] didn't even use the term "blockchain", preferring instead to use the term "distributed ledger technology" - I wonder if they are concerned about negative connotations the term "blockchain" might be starting to have.

[0] https://www.bankofengland.co.uk/-/media/boe/files/payments/r...

I do agree with you on a lot of your points, the current state (and "value") of cryptocurrencies is extremely over inflated, but I don't think we can use the dotcom bubble as any sort of guide to when the blockchain bubble will burst. They are fundamentally very different things.

I feel like most cryptocurrencies will die a slow death over the coming months with the ones that actually provide value decreasing in price but not dying, with the herd will be culled down to ~20 or so. Blockchain is a useful concept, and tokens of value are a key concept in adoption.

I don't think we've seen any truly useful applications of the technology, but I also feel like its too early to write it off completely. It would only take one great application of blockchain for it to become an important part of our daily lives, but I think the focus at the moment is in the wrong areas.

You might like saito.tech. Very early stage, but it is a genuine solution to blockchain scaling that powers applications rather than apps. Be sure to check out the decentralized (on-chain) Reddit for a sense of how flexible a big-data PKI network can be.

There are other interesting projects nearing fruition like Radix and Gun that are trying to create different types of distributed databases that have some of the properties of blockchains.

Blockchain tech is still 10+ years from being practical. But it could be useful in developing countries with bad governments and endemic corruption. In the 1st world things work well enough.

The dot com era actually started in the late 1960s with packet switching networks. It took 50 years to get mass adoption. The first really working cryptocurrency was in 2008. I expect mass adoption by 2058.

Web != packet switched network. But, if you want to go that far, the first cryptographic hash functions were being designed in the late 1970s, so cryptocurrency has a decade or so to go before it'll be "late" by that standard, and it's progress so far has been lackluster compared to the internet's as of 40 years after it's "packet switching network ancestry".

The web is an application. PSNs are a fundamental innovation. A fundamental innovation is a conceptual leap that brings about a qualitative (not quantitative) jump in capability. Before it something is simply impossible. After it that thing is possible.

The fundamental innovation was the development of the very first practical decentralized network that can achieve secure reliable consensus among minimally trusted peers. Bitcoin's way is not very efficient and may not be the only way to do it, but it proves that it's possible and shows one way it can be done.

It generally takes an absolute minimum of twenty to thirty years for something to go from lab bench prototype to mass adoption, and that's fast. 50 is more typical. Bitcoin was a lab bench prototype.

But what about the bubble? Cryptocurrency is indeed in a bubble, but it's a bubble built on radically underestimating the amount of time (and work) it will take to go from proof of concept to mass adoption. As the saying goes people tend to over-estimate progress in the near term and under-estimate it over longer spans of time.

The cryptocurrency bubble is also not very big. The numbers are huge from the perspective of an individual or a small company but from the perspective of the global economy it's the equivalent of taking a piss in the ocean. If you look at the global financial industry as the target market, cryptocurrency collectively hasn't scratched the surface. Far more money than the entire cryptocurrency valuation crosses the Atlantic and Pacific in the form of conventional wire transfers every day.

Reply on news.ycombinator.com