Hacker News new | past | comments | ask | show | jobs | submit login
Ask HN: What are new applications that can be built on blockchains?
90 points by noloblo on June 17, 2017 | hide | past | favorite | 64 comments
i am not as interested in the price volatility of this asset class more inspired by the crypto anarchism that they enable and new trustless, non custodial and truly decentralized apps that can be built on top of modern blockchains

I think it makes a lot of sense to build a DNS system using contacts. Names can be auctioned off and given automatically to the highest bidder. It is also possible to set up a contract where you could trustlessly sell names. "Send $200 to an address and you will get the name". There is an auction like this in progress already, and you can participate now if you like http://ens.domains

You could do even more interesting things trustlessly. For example say, you own a valuable domain. You could borrow money and use the domain as collateral. If you do pay back, or stop making monthly payments the domain is automatically and trustlessly transferred to the lender. Or it can be put up for a trustless public auction with the proceeds going to lender to repay the loan and the rest to you, the ex-owner of the domain who failed to pay back his loan.

Who is being trusted to accurately report whether the borrower has paid? I have never understood how the "oracle" problem gets worked out in practice.

The amazing this about this particular application is that an oracle is not required at all. So it's completely automatic and trustless. The contract (think of it as a class in Java) has a single function which looks something like this: function acquire_name() { if (msg.value > $200) {self.registrar.transfer_ownership(self.domain_name, msg.sender);}}

registrar is a pointer to an instance of the domain registrar that keeps track of all names. Here's an actual implementation of the registrar: https://github.com/ethereum/ens/blob/master/contracts/HashRe...

And how would this help ? The economical problems people face in this day and age are not, at all, that contracts aren't enforced. Quite the contrary.

The opposite is true: the problems people face is that the only contracts available to them, from rents to cell service, have onerous and unfair terms, favoring whichever side of the contract requires investment, and this unfairness goes up to a ridiculous extent.

The last thing we need is more thorough enforcement of them. Hell, I do believe even the owners themselves would demand the ability to "do business". For instance, to let someone hang on to a name and remain in business while they search for a way to repay their debts.

I am not sure what you're asking when you say "how would this help?". Do you not see any usefulness at all in people being able to sell domain names to each other trustlessly without needing to involve a third party?

I do see the usefulness of that approach, but I also see an amount of risk. A disaster waiting to happen.

Let's say, for instance childpornhere.eth goes live, someone pays and receives mail with pictures of a kidnapped child. Havethisguykilled.eth. Sponsorkillingblacks.eth. tipISISnottoattackyourkids.eth. And so on and so forth.

The bitcoin "hostage" situations that we see now. Famously the encryption viruses, but the same is happening with cloud credentials (e.g. gmail or hotmail accounts used for businesses. Passwords get changed, last mail is "if you want them back pay $1k to <bitcoin address>")

Second, securities fraud. If anything was learned from bitcoin in finance, it's that we cannot live without exchange laws, which bitcoin exchanges did not have. Tradehill, Mt. Gox, Bitcoin central, Bitstake, BitSpark, all and many have gone bankrupt, stealing large investments in the process through outright fraud (TradeHill, incidentally the guy who committed that fraud has a new bitcoin exchange), or through incompetence and getting hacked (which of course could always be just straight up fraud, I mean does anyone buy the Mt. Gox story, for instance ?)

Second, addendum: outright general securities fraud. E.g. sudden changes in liquidity that I'm sure in most cases are mere accidents (well, 2 years ago in most cases, today, I'm not so sure), but cause short-term very quick value changes that even in the accidental cases most people find themselves victims to. More generally, the sale and trade of bitcoin derivatives and the practice of the issuers of said products to invest heavily for short term price changes. Or, causing an artificial and massive drop in liquidity through ddossing an exchange (hell, I've worked for actual financial exchanges and seen people pull this one off, but in bitcoin it's commonplace and unpunished)

Thirdly, tax evasion. People use bitcoin, and eth, to avoid paying 0.2% or so commission to banks. We are to trust that these same (very rich) people have no issues paying 30% or so in tax. Riiiiiiiiiiiiiiiiiiiiiiiight.

Do you really want to allow that ? This is (mostly) what is enabled here.

In many (most) parts of the world the legal system for enforcing contracts isn't all that reliable and trustworthy.

An Etheruem-based contract is a big improvement over a traditional contract enforceable only by a slow, corrupt court.

Pretty much all technology increases the surface area for fraud. However, the benefits usually outweigh that increase. Blockchains are no different. But the benefit is likely to accrue more to people in less developed countries.

Okay ... let me put it this way. 4 years back I transferred a number of bitcoin, in the low 3 digits, to tradehill, in order to explore this whole finance thing and trading.

We all know how tradehill, and mt. Gox, and ... ended, so can you explain to me how this works better without the option to go to a court ?

Sure, don't use shady exchanges.

Using a regular old currency you could end up being ripped off by some fraudster, take them to court and lose the case. The non-cryptocurrency world isn't 100% safe and guaranteed to be protect you from fraud either.

It would be foolish to suggest that there will never be any fraud in the blockchain world. The risks are just different. For some people in some places the risks are probably a lot higher than the fiat currency status quo. But there are also plenty of people in places where enforcing a contract by code could be a lot less risky and a lot more appealing.

Also, I should add, just to clarify, I think we're watching the very early days of blockchains. Things are still pretty rough. I'm not especially optimistic about any one of the current crop of cryptocurrencies. But I am optimistic about the underlying ideas over the next, say, five to twenty years.

> Sure, don't use shady exchanges.

Neither TradeHill nor Mt. Gox were considered shady exchanges, at any point when they were used. But you do realize that the standards for this are somewhat low.

E.g. both the owner of bloomberg.com and the owner of businessinsider.com are convicted of securities fraud (in fact businessinsider.com was extremely likely started exactly because Henry Blodget was convicted of fraud). At the time, the owner of Tradehill was not convicted of fraud, and while he is now, and yet he's still running a bitcoin exchange.

However, the people defrauded by both of these individuals in "fiat-world" saw 80%+ (Blodget) or 100% (Bloomberg) of their money back. As far as I know, nobody saw their Tradehill money back.

> It would be foolish to suggest that there will never be any fraud in the blockchain world.

Are you seriously naive or misinformed enough to believe this ? Fraud, law breaking and tax evasion are the main, almost the only, reason bitcoin is being used at all. You have heard of Silk Road, right ?

I've seen numbers suggesting that Silk Road was responsible for 70%+ of all bitcoin transaction at one time, and this isn't even counting the many other kinds of fraud and crime, like data ransoms, actual ransoms, money laundering (esp. out of China, in fact most miners are money laundering operations according to many sources)

I don't really follow. Is this transaction in eth or dollars? If the former, how does the algorithm know what the exchange rate is? If the latter, how does it know a USD transaction has even taken place?

It sounds like you're talking about a hypothetical future in which eth is just as useful as USD, meaning that the risk in holding eth and converting it to USD is negligible, and that the contract writer doesn't have to worry about the exchange rate changing much. If so, then it seems like you're sort of solving the easy part and skipping over the hard part.

How would ETH name service Domains be integrated with the browser or this is supposed to work only in onion Tor and Mist browsers

Update fixed formatting

FYI- while cool looking, wrapping your entire multiline comment in code tags makes it fairly unreadable on mobile (only four words show at a time).

There is a Chrome extension that does this for you for IP addresses (I think). All ethereum wallets are also implementing support. You can make your Ethereum domain name point to anything you want. So instead of sending your transaction to a long hex string, you can send it to "johndoe.eth".

Thanks but chrome and electron-based applications really heat the Mac, get the fan running all the time and bringing it to a halt.

Are there projects to build native cocoa apps on top of eth/btc much like aquamacs for emacs

Democratic Autonomous Organizations -- which are corporations whose law is created not by human organizations; the law for ricardian contracts is instead defined by a relationship between trustless consensus and formally verifiable logic.

See the og tao here: http://chriseth.github.io/browser-solidity/?gist=192371538cf...

and read about the $50 million tao debacle on wiki: https://en.wikipedia.org/wiki/The_DAO_(organization)

more about ricardian contracts on ethereum here: http://iang.org/papers/intersection_ricardian_smart.html

That's "Decentralized Autonomous Organization".


Why was the DAO abandoned? It seems like someone could just fix the bug, and run the sale again, expecting a lower valuation.

The idea is still good, right? A bug shouldn't end the whole project.

There are many teams building projects similar to the DAO right now (except much more thought out), for example https://www.reddit.com/r/MakerDAO/. In the future, we will definitely see more DAOs - read this article if you're interested https://blog.ethereum.org/2014/05/06/daos-dacs-das-and-more-...

The original DAO was just a preview and it was too early when the whole Ethereum ecosystem was young (and still is, it's less than 2 years old).

The idea was that code-as-contract was clear, unambiguous to all participants, and obviated the need for human intervention to enforce correctly.

You forget to tell what happened after that.

Someone someone found contract terms (relating to starting a new sub-entity to the DAO) that favored them, started transferring the eth to themselves using this method (they "probably" gained the ability to dictate how the eth would be spent after an initial period, which plenty of people are desperate to point out isn't the same as owning it). And then ...

Humans intervened to have the terms enforced "correctly" (this was regarded as a hack). They even rolled back quite a few normal transactions that had nothing to do with the attack (because just like all other central banks, the ethereum company cannot risk/be bothered to figure out what transactions are about and come up with a solution that affected just the hack and nothing else)

I imagine this will happen again, and again, and again. This is very clear to anyone who's done the least bit of law : humans cannot be trusted (not even by themselves) to write rules and then uphold them "under any circumstances". We must have governments and a legal system to prevent disaster. The easy tricks I see work is that humans can loan eth from others, then have multiple accounts, only one of which is subject to confiscation. Double collateralizing (pledging the same asset for multiple loans, then disappearing with the loans, recently popular in China).


Just for the record, no transactions were rolled back. A special one time transaction was generated that moved all the ether from the Dao contact to a withdrawal contract, from which all investors could get their ether back. Majority of the users upgraded their clients to be on the chain with this special transaction. The rest of the people (~15%) remained on the original chain without the special transaction, that chain is now called "Ethereum Classic".

The DAO was unwound, including everything it had done. If that's your idea of "no transactions rolled back", I submit that that's exactly the sort of distinction without a difference I was talking about.

This would be akin to solving a bank heist by simply rolling back all transactions involving that bank, including 2nd and 3rd degree transactions to before the heist.

So if some person sold a house to a buyer with a mortgage from the bank close enough to the bank heist, and then went out and bought a bottle of champagne to celebrate with the buyer, the bartender would wake up the next day to find the payment for the bottle missing, and the house would be unsold, just so the investors in the bank would be guaranteed the money they invested to be loaned out.

I mean I understand. This simple and automatically enforced contracts is every loaner and landlord's wet dream. They get that the rules are stacked against the tenants. They can rob tenants blind using this game. They don't realize that they cannot play this game against the banks, the millionaires and/or billionaires of our little planet, and they don't know any other game to make money.

Distributed ledgers [0], if you're a Fortune 500 [1] [2]

[0] https://en.wikipedia.org/wiki/Distributed_ledger

[1] http://news.sap.com/sapphire-now-sap-cloud-platform-blockcha...

[2] http://www-935.ibm.com/industries/retail/supply-chain/

The Linux Foundation's related project:


Turn your brain's marketing parser on before attempting to read any of this stuff.

Get rid of ticket scalping.

If tickets get passed on from the buyer you pay a extra fee ( which increases with each hop ). On ticket return, a pool will be created where you can bid on -> sold tickets above the original price are going to be splitpaid to the venue and the artist / ...

Requires a solution to the identity problem otherwise scalpers just sell off whatever nonce they purchase the tickets under

you are right. Just enter your full name after you hit the buy button.

You and I have very different definitions of the word "solution"

Its a simple fix. The venue has to control the name on your id with your ticket name. It has to be personalized in some way and it seems reasonable to keep your "real ID" away from the blockchain account. Even though the venue can identify your wallet on entrance... so there is a small design flaw :)


We are working on electricity trade using blockchain, enabling renewable energy generator to trade with neighbours in local grid.

That sounds interesting is there a link to try the prototype who is we

Some obvious ones:

- Cloud commodity (storage, computing - think of decentralized AWS)

- Identity & reputation systems (think of global credit rating & KYC)

- One to one services cutting the middleman (consultation, tutoring, renting)

- Autonomous agents self negotiating contracts (common example is your car paying for its parking)

Sorry for being a noob, but how is something like "decentralized AWS" enabled by blockchain technology?

Check out https://golem.network or http://sia.tech You rent out your unused CPU (Golem) or unused disk space (Sia) and get automatically rewarded

Still it's early days, but watch these talks on IPFS[1] and Golem[2] to get a basic idea.

[1] https://www.youtube.com/watch?v=HUVmypx9HGI

[2] https://www.youtube.com/watch?v=DGinj5K3tcI

You can do immediate payment - and with this, hard spending limits.

One of the things about blockchain is that it has a history, that history makes it to some extent atomic.

This means it can be used to turn digital assets which are normally considered infinite in that it can be copied and pasted, into scarce resources.

One example could be a 2nd hand ebook market where books could actually increase in value because it's been owned by this and this person.

Check out the top coins here: https://coinmarketcap.com

From there you can click on any coin and get a link back to each project's website. Exploring those should cover a good chunk of the use-cases that are popular right now.

That's a good start.

I'm not looking for the use cases of the coins I'm looking for applications that can be built on top of the Blockchain enabled by the new cryptographic go i'm not looking for the use cases of the Coins I'm looking for applications that can be built on top of the Blockchain enabled by the new cryptographic coins - what are the non-obvious applications

How about a simple clock in clock out system for payroll. here is a directory of dapps https://dapps.ethercasts.com/

Honest question: what are the deficiencies of current clock-in / clock-out systems that blockchain technology might help solve?

Not GP, but current systems usually are not manipulation proof. Using a blockchain makes post-clock-out modifications visible immediately.

Sure, but what I'm curious about is: is that a real problem in practice that organizations have been fighting, or is it just a problem in theory? If I told an organization, "I've built a blockchain based time tracking system that is 100% impossible to be tampered with", would they say, "oh thank goodness this is a huge problem for us!", or would they say, "oh that's sort of neat but our tampering related loss is already within acceptable bounds"? My sense is that it would be the latter.

It is not only about being impossible being tampered with. It is about trusting the system. Since we trust the blockchain with 100 + billion dollars. trusting it for a time sheet system should not be that hard

Sure, but I think you're missing my point. Is this a problem that really needs a solution or is it a solution in search of a problem? Are time tracking systems not trustworthy enough already? Perhaps I'm just ignorant, but I've never heard of this issue, and it just sounds like the sort of thing that sounds like a problem to techies, but isn't really a problem in practice.

Timesheet cheatery is an extreme problem - often enough timesheets are... adjusted afterwards to hide violations of labor law or simply to dock pay.

Interesting! Thanks for the info.

You can multisignature contracts with advanced options for payment or you can gamble on predictions on future for example, both in trustless and decentralized way

Trust less Decentralized exchange

Open alternative to Uber, lyft and Airbnb

Isn't the trust in those services more at the physical world level than the technical level?

Yes, but the idea is now we can remove the custodians out of the mix. Why pay airbnb? The networking, booking, hopefully identity verification can all be done p2p

I think a big problem with these services is that your rating belongs to Uber instead of the user. It would be nice to put your reputation from all the services you use (Uber, Airbnb, reddit karma, eBay, credit rating, etc) in one central place that is not controlled by a single corporation.

The dream it's to then be able to share a portion of your information with another party without revealing everything. For example, you can share "my credit rating is above 750 and I'm 32 years old" and nothing else. This can be accomplished with zero knowledge proofs which are going to be possible on Ethereum later this year.

Thanks, that makes a lot of sense.

Edit: There does seem to be a bootstrapping issue though. Right now the centralized entity is basically vouching for new people until their individual reputation builds.

Google golem dapps, very interesting use case.

Spelling and grammar checkers.

Would it work in a distributed context

And punctuation checkers.

far more accurate click trackers


Do you mean the elections or prediction markets


Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact