4 ms·
>>So aside from coins, where else is it that you would not trust a central database, but you can verify/trust the entries being added to that database? I would
by CryptoPunk 5y ago
>>So aside from coins, where else is it that you would not trust a central database, but you can verify/trust the entries being added to that database?
I would not trust a central database for any application, even those where I have to trust the verifiers of the entries to that database.
E.g. I would prefer stablecoins over credit in a bank.
With a decentralized database, immutability and permissionlessness are the default, until the trusted third party actively intervenes to strip you of these privileges.
With a centralized database, access is by default denied, and requires active intervention from the trusted third party, in the form of a grant of permission, to acquire.
- yashg 5y agoYou can't trust a central database but you can trust a central blockchain? A proof of work blockchain which is trustless and permissionless and has many nodes can be trusted to be tamperproof and immutable, but it is also slow and inefficient (thanks to PoW). A PoW blockchain is inefficient by deaign, it's a feature not a bug beucase that is what gives it's trust without trusting. But when you make that blockchain private, trustful, and permissioned (only designated nodes can add blocks) you are taking away the very properties that make the blockchain immutable. And it has to be done because that is what enterprises would want. So an enterprise blockchain like Corda/Hyperledger is no different than a central database. And no serious business least of all banks would do their business on a public truly distributed open blockchain.
- CryptoPunk 5y agoI never suggested I would trust a centralized blockchain.
- throwawaylinux 5y ago> I would not trust a central database for any application, even those where I have to trust the verifiers of the entries to that database. I don't see how that's reasonable. What am I missing, surely virtually nobody would be willing to spend the massive resources of a blockchain for that kind of risk model. > E.g. I would prefer stablecoins over credit in a bank. So aside from coins, what's another concrete example? > With a decentralized database, immutability and permissionlessness are the default, until the trusted third party actively intervenes to strip you of these privileges. > With a centralized database, access is by default denied, and requires active intervention from the trusted third party, in the form of a grant of permission, to acquire. This sounds very much like the buzzword snake oil I've heard many times before, but I'm willing to be open minded about your non-coin example.
- CryptoPunk 5y ago>>So aside from coins, what's another concrete example? What you wrote earlier is: >>Crypto coins are unique in that you can verify them mathematically. They have no connection to anything else out in the real world so they don't have that problem. Stablecoins cannot be verified entirely mathematically, as they have a connection to real world bank notes, so my example doesn't fit your exception. >>This sounds very much like the buzzword snake oil I've heard many times before, but I'm willing to be open minded about your non-coin example. I'll use the coin example, because it's easy to demonstrate the principle. I can move my stablecoins, without having to ask a third party custodian for permission. With bank credit, I need to ask the bank for permission to withdraw it, or transfer it.
- throwawaylinux 5y agoThe coin certainly can be verified, otherwise you couldn't move it. A stable coin where you trust whoever is providing the backing for that asset is no different from a crypto currency in that way, the trust model still requires that you trust someone to honor the coin or value it in some way. So it's a bad example. The interesting part about it that you can move coins between people who don't trust one another without a 3rd party is entirely due to the remotely verifiable nature of the thing, like any other coin. That is the one and only interesting thing about it, and that's common to all these kind of coins. I'm asking for something else.
- CryptoPunk 5y ago>>A stable coin where you trust whoever is providing the backing for that asset is no different from a crypto currency in that way, the trust model still requires that you trust someone to honor the coin or value it in some way. I don't follow. With a cryptocurrency you don't have to trust any one. The cryptocurrency can't be redeemed for anything, and thus there's no one to trust for honoring the claim upon redemption. >>The interesting part about it that you can move coins between people who don't trust one another without a 3rd party is entirely due to the remotely verifiable nature of the thing, like any other coin. Yes true. And that applies to any digital property on the blockchain. The challenge for digital property that is a claim on a real world asset is being able trust a third party to honor the claim upon redemption of the digital property. Some parties have solved that for claims on national currency, and thus we have a few stablecoins. Whether this model can extend to real estate, automobiles, commodities and other assets/goods remains to be seen.
- rank0 5y agoYou realize all this web3 stuff runs on top of the internet right? There are centralized CAs, domain/ip registrars, ISPs…shit even the libraries that comprise your entire computing stack all come from centralized databases! Your comment is stored on a database. USDC/Tether have private databases and their collateral is held by third parties in again their own database.
- CryptoPunk 5y agoI would prefer all of the databases used within the internet infrastructure to be immutable append-only blockchains maintained without any centralized trusted third party.