Hacker News Reader

Comment by Moodles | Hacker News Reader

Still waiting for a use case of blockchain outside cryptocurrency also.

There isn't. If there are, none of them would be viable for-profit business entities like they're claiming. I can see a nameserver (like DNS) using a blockchain. Or maybe a distributed identity service can use a blockchain. These would have to function as public goods owned by no one; otherwise why not use existing DNS or existing government IDs. This pretty much rules out any mention of "blockchain" in a for-profit company's pitch deck as ever being useful to the world.

DNS on a blockchain is actually the very first altcoin, before even Litecoin. 2011.

It’s not really usable though; all the “dns on blockchain” means just “well we can store arbitrary data on the blockchain, some are parseable as DNS records maybe, and then I guess we can parse all the chain to get all the records???”

it was as meaningful in 2011 as it is in 2020

A blockchain alone is a horrible data structure for random access lookup. There were so many hacks that had to be done to make Bitcoin work efficiently on BerkeleyDB and LevelDB.

Ethereum lets you do general purpose computing, so there are nameservers that just use a hash table. That's how ENS (.eth) works. But at that point it's not really "DNS as a blockchain"; it's more like ICANN but running on Ethereum miners's computers instead of 8.8.8.8.

You are making an assumption about how blockchains work. Do you think the linked-list like structure of how blocks are added to the chain is how data is stored and retrieved?

For reference: https://en.bitcoin.it/wiki/Bitcoin_Core_0.11_(ch_2):_Data_St... https://hackernoon.com/getting-deep-into-ethereum-how-data-i...

I'm aware it is implemented more efficiently. The problem it is trying to solve is orthogonal to what DNS does: search vs. verification.

DNS needs to resolve an name on the order of nanoseconds across a distributed system. It is designed with this in mind. A blockchain is designed with integrity in mind. Wallets can verify transactions on the order of nanoseconds with SPV--also distributed because a client can get the intermediate block headers from any server. Querying data, like for DNS, has constraints different from verifying integrity, like for a ledger.

Sure, you can use a blockchain for storage, but who do you trust to do the lookups? It can't be on chain because it will be far too slow. You'd end up having to trust some entity like Cloudflare to run a DNS-on-blockchain node that stores the ground truth, going full circle back to regular DNS. What's the equivalent of SPV for DNS-on-blockchain where I can do and verify name resolution with limited resources?

In terms of access time, the standard would be O(1) for hashtables which would be my best guess at how to best represent an ip to domain mapping. Reading from on chain data is done locally (if you have your own node) and is near instant, but for wider reach I think DNS over HTTP would be needed or wasm clients that can be run in a browser.

Here is a possible implementation in Ethereum: https://medium.com/the-ethereum-name-service/ethdns-9d56298f...

Consider a timestamping service that's based on mathematical proof instead of social proof.[0]

But I agree, most often blockchain outside of a cryptocurrency doesn't make sense. This use case also depends on the security of an existing cryptocurrency.

[0] https://whycryptocurrencies.com/timestamping_service.html

Decentralized timestamping is literally what Bitcoin is, there's no other way of describing it. Ordering transactions canonically is timestamping.

My point is that you can leverage this by inserting extra data on a transaction, so you can timestamp any message.

A cryptocurrency is basically a timestamping service. This was mentioned in the original whitepaper.

> The solution we propose begins with a timestamp server. A timestamp server works by taking a hash of a block of items to be timestamped and widely publishing the hash, such as in a newspaper or Usenet post [2-5]. The timestamp proves that the data must have existed at the time, obviously, in order to get into the hash. Each timestamp includes the previous timestamp in its hash, forming a chain, with each additional timestamp reinforcing the ones before it.

https://bitcoin.org/bitcoin.pdf

> Still waiting for a use case of blockchain outside cryptocurrency also.

https://news.lockheedmartin.com/2017-04-27-Lockheed-Martin-C...

https://www.ericsson.com/en/press-releases/2014/9/ericsson-a...

https://nortal.com/blog/blockchain-healthcare-estonia/

etc.

The history of a second hand car is currently scattered all over the place, even with missing data. Some entities try to collect it all, but it is difficult.

Every car has a unique VIN, so imagine all data of the car is written to a blockchain. Maintenance, mileage, accidents, etc. When you buy a car, you check the blockchain and get a complete history, unable to tamper with afterwards.

Of course you have to trust the entities that write the info. But the more trustworthy entities, the harder a malicious entity can withhold or provide false info.

Sort of a silly assertion as what's baked into your question is.

"I'm waiting for a use case of blockchain outside assets"

Which there are plenty of entities using or applying blockchain tech as a means of tracking the immutable transfer of assets. That's a fairly large market place. Does it need to be more than that?

This shit isn't gonna cure cancer, tell you why you sometimes you have wondering sexual thoughts, or run your life for you.

It may be valuable for some applications.

The SAFE networks plans to use a Blockchain to create a fully decentralised internet storage solution seems like a reasonably good use case.

https://safenetforum.org/

git's a blockchain. you have commits (blocks), which contain a hash of the previous commit (chain).

I think your analogy works better if branches are chains and commits are transactions.

Blockchain (as in a Merkle tree with some way to decide on a particular path to a leaf as the canonical "correct" chain) is useful in plenty of things, eg git.

Blockchain as in the above with a decentralized decision function (proof-of-work, maybe proof-of-stake) is less likely to be useful outside cryptocurrency.

Typically when people talk about "blockchain" they mean the latter.

Decentralized monetization of data and media. F* central brokers like facebook.

Identity looks like a potentially interesting use case.

https://www.computerworld.com/article/3244128/how-blockchain...

This totally collapses for most real world uses of identity. If it's used for anything important then you need to be able to recover your identity if your house burns down with your keys inside or you lose your passphrase due to brain injury (or you just forget). You need centralization for most real-world uses of identity.

Could you have an identity where if the above happens you just start over? Yes, as long as it's not "too" expensive. But if you make it too cheap you risk Sybil attacks. So your identity system can't really be used for reputation. What's left? Not much IMO.

Timestamping use cases like 'I created this work of science/art at this time and I can prove it' are good, especially if you broker with an agent that can reliably and securely store keys. Identity is hard IMO.

> Could you have an identity where if the above happens you just start over?

Reading the W3 spec for Decentralised Identifiers is probably a bit dry, suggest looking into other sources on this matter.

Regardless the answer to your question is yes.

A lot of smart people have been working on this very thing for years.

https://www.w3.org/TR/did-core/

decentralized cloud storage and hosting

Synaptic Health Alliance

https://www.synaptichealthalliance.com/

Reply on news.ycombinator.com