6 ms·

I feel the IPFS blog post gave me a glimpse of how blockchain can be cool. In certain cases, the need for decentralization is so strong that it makes sense to b

I feel the IPFS blog post gave me a glimpse of how blockchain can be cool. In certain cases, the need for decentralization is so strong that it makes sense to build a blockchain system for it (or other secure P2P network thingy). For example, authentication/identity using blockchain technology seems interesting to me. [1] https://blog.ipfs.io/decentralizing-the-internet-s-root/ https://blog.ipfs.io/decentralizing-the-internet-s-root/

wyager 5y ago
The only two known usage of blockchain that actually seem to offer a clear advantage over some other strategy (running the thing centralized, not having coordination at all, etc) are A) solving double spend (a la Bitcoin) B) solving zooko’s triangle (a la Namecoin, or the clone you mentioned from ipfs). Identity on blockchain beyond resolving zooko’s triangle (I.e. stably mapping pseudonyms to pubkeys) typically does not offer an advantage over (partially) centralized identity providers. E.g. putting anything related to government ID on a blockchain is pointless because you might as well have the ID issuer run a central or federated service.
- karpierz 5y ago
  Why does Bitcoin have a clear advantage in solving double spend over a centralized solution?
  - krageon 5y ago
    > centralized If you want to trust a conglomerate of folks who have a vested interest in keeping you small and poor to solve your problems, then there is no advantage. If you do not, then the advantage is exactly that it is not centralised.
    - acomms 5y ago
      Why are you so convinced that cryptocurrency solves the "outside forces much larger and more powerful than me have influence over the system" problem? It exchanged a solution to some problems for a new set of problems to be solved.
      - krageon 5y ago
        You are questioning what I'm saying by telling me something incredibly vague. I'm afraid I just don't know what to do with your question.
        secondaryacct 5y ago
        I can rephrase it: why do you think it impossible for vastly resourceful entities to acquire so much power they can control the decentralized system - in a softer way, probably but still. Imagine a world tmr where bitcoin is everywhere. You ll have mining conglomerates in the cheapest electricity providers working as cartels to help each other and protect their interests (so can envision centralizing a blacklist of addresses they dislike, coordination on software version etc), de facto cartels of whales (like todays "billionaires" people describe as a collaborative force), and large lenders who concentrate capital and therefore still can buy out whatever they want (today called banks). How do you think the blockchain solve any of that?
        megameter 5y ago
        It's easy. Copy-paste the blockchain and start over with a fork. When the system delegitimizes itself, a debt jubilee lies around the corner. This stuff is just glue now - a cheap way of making a financial system.
        DangitBobby 5y ago
        So if you don't like the history of the currency, scrap it entirely and start over. That seems like the opposite of the point of Blockchain.
    - imtringued 5y ago
      What keeps people small and poor is the insistence that people should save in monetary terms. It's not rocket science. When you keep your wealth in monetary terms, you don't get to have it in physical terms. Money represents how much real wealth you could have if you spent it. It's deferred consumption. There is no real wealth in something that didn't happen.
      - Cipater 5y ago
        Money is a store of value. Money is a measure of wealth.
  - mcherm 5y ago
    It does not. Bitcoin (and it's successors) are a decentralized solution to the double spend problem. If "decentralized" is a requirement, then a cryptocurrency can be an excellent solution. Other known solutions (like "just trust everyone") don't scale well (an understatement!). If "decentralized" is not a requirement then there are other solutions to the double-spend problem.
    - mdoms 5y ago
      So it's your contention that the current traditional financial system "don't scale well (an understatement!)"?
  - snarf21 5y ago
    Because the trust is distributed. One single person can't cause the entire system to break down or steal all them money. With blockchain coins, the incentives are aligned. The weird thing that happened with BTC that I don't think Satoshi could forsee is that the more valuable it got, the less usable it got. This is one of the unsolved issues: how to create decentralized consensus very very quickly that can't also be attacked via a PoW that is too easy.
    - ironSkillet 5y ago
      Do you know of any lines of research that are out to prove mathematically that it is impossible to have these desirable characteristics simultaneously? I.e. any improvement in consensus efficiency necessarily gives up some amount of decentralization, appropriately defined.
      - ItsMonkk 5y ago
        This is the sort of thing that really needs to be questioned. 1. How much security do we need for this transaction? 2. For this block of transactions? 3. Can we have varying levels of security? What we are ultimately trying to know is, is it profitable to double spend? If the answer is yes, then no one should be transacting, and anything they do transact is a gamble. If the answer is no, there's much less risk, and the farther the answer is no, the safer the transaction. But we likely don't need as much security as we do now. In fact, it's very easy to handle at a user level. A. Put the transaction on the blockchain. B. Wait for additional blocks. C. If your transaction is huge, wait for more additional blocks. Wait a week, a month. This is an important transaction. When the blockchain has not reverted for such a long time, finally do the transaction in the real world. If you are transacting with yourself, you don't need any security at all. If you are transacting with a bar and the bar knows you, much like a tab, you don't need much security. If you are transacting with an anonymous person, you want high security. So security is currently handled on a blockchain by blockchain basis, where it should be a transaction by transaction basis. What if you can't wait a month? Then you should put heavy fees on that transaction and it should be put in with other higher fee transactions. 4. How does security change with transaction size? Does it make sense to have limits? 5. How does security change with total transaction block size? Does it make sense to have limits? 6. When the coin price goes up, how does that change these factors? 7. When the mining reward in coins drops, how does that affect security? 8. When Bitcoin, in 100 years, moves to 100% transaction fee system, what happens to security? What would the transactions fees look like? 9. In 50 years when the block reward has been halved 5 times, either the price of Bitcoin moves to $1 million, or the security drops. Is that okay?
  - wyager 5y ago
    If you’re fine with all the downsides of trusting you bank, the bank’s central bank and currency issued, the bank’s government, etc then that obviously makes Bitcoin less appealing for you. Remaining upside includes reduced friction for things like international settlement. Don’t tell me that international settlement is easy with traditional banks - I am a financial system power user and it is not.
- cryptica 5y ago
  I think this is an accurate assessment but I think authentication and account management should be seen as separate from 'identity'. Authentication can be useful without being associated with a real-world identity; for example, it can be useful for spam-prevention as an alternative to signups with captcha. You could theoretically log into multiple services which you don't trust by signing messages using your private key without revealing your passphrase to any of those services.
cryptica 5y ago
Authentication using the blockchain makes a great AI-resistant alternative to signups with captcha. The same account could be used by multiple independent services and this removes the need for a signup process altogether. It also removes the need for password management since you can use a single account/passphrase to log into various services without revealing your passphrase to any of these services; you can simply sign a message to prove to any service that you control an account (and know the passphrase) without actually revealing the passphrase to any of the third party services.
- UncleMeat 5y ago
  > The same account could be used by multiple independent services and this removes the need for a signup process altogether. This is only marginally better than SSO with Google/Apple/Facebook/whatever. It dodges the "Google closed my account" problem, which is a meaningful improvement. But I'm not fully certain why you need the public key to be part of a blockchain for this to work. Couldn't you just publish your public key identity on any number of considerably less wasteful systems? As far as I can tell, the only advantage here is the mapping between the name "UncleMeat" and my public key would be able to be reused across services... but in so many cases I don't want my identity to be consistent across services.
  - cryptica 5y ago
    - Resistant to centralized censorship (as you pointed out) - It allows you to use the same passphrase for all services without compromising security since your passphrase never needs to leave your own machine; it is never sent over the wire. You just send signatures to different services. - The cost associated with purchasing tokens needed to initialize an account on the blockchain would serve as a spam prevention mechanism; an alternative to SIM cards which most centralized services rely on today as a cost barrier to limit the creation of spam accounts. - Superior integration potential between different services/systems provided by different companies/groups since they can all refer to the exact same account on the same blockchain and provide new ways to unify data between the different services.

wyager 5y ago

The only two known usage of blockchain that actually seem to offer a clear advantage over some other strategy (running the thing centralized, not having coordination at all, etc) are A) solving double spend (a la Bitcoin) B) solving zooko’s triangle (a la Namecoin, or the clone you mentioned from ipfs). Identity on blockchain beyond resolving zooko’s triangle (I.e. stably mapping pseudonyms to pubkeys) typically does not offer an advantage over (partially) centralized identity providers. E.g. putting anything related to government ID on a blockchain is pointless because you might as well have the ID issuer run a central or federated service.

karpierz 5y ago
Why does Bitcoin have a clear advantage in solving double spend over a centralized solution?
- krageon 5y ago
  > centralized If you want to trust a conglomerate of folks who have a vested interest in keeping you small and poor to solve your problems, then there is no advantage. If you do not, then the advantage is exactly that it is not centralised.
  - acomms 5y ago
    Why are you so convinced that cryptocurrency solves the "outside forces much larger and more powerful than me have influence over the system" problem? It exchanged a solution to some problems for a new set of problems to be solved.
    - krageon 5y ago
      You are questioning what I'm saying by telling me something incredibly vague. I'm afraid I just don't know what to do with your question.
      - secondaryacct 5y ago
        I can rephrase it: why do you think it impossible for vastly resourceful entities to acquire so much power they can control the decentralized system - in a softer way, probably but still. Imagine a world tmr where bitcoin is everywhere. You ll have mining conglomerates in the cheapest electricity providers working as cartels to help each other and protect their interests (so can envision centralizing a blacklist of addresses they dislike, coordination on software version etc), de facto cartels of whales (like todays "billionaires" people describe as a collaborative force), and large lenders who concentrate capital and therefore still can buy out whatever they want (today called banks). How do you think the blockchain solve any of that?
        megameter 5y ago
        It's easy. Copy-paste the blockchain and start over with a fork. When the system delegitimizes itself, a debt jubilee lies around the corner. This stuff is just glue now - a cheap way of making a financial system.
        DangitBobby 5y ago
        So if you don't like the history of the currency, scrap it entirely and start over. That seems like the opposite of the point of Blockchain.
  - imtringued 5y ago
    What keeps people small and poor is the insistence that people should save in monetary terms. It's not rocket science. When you keep your wealth in monetary terms, you don't get to have it in physical terms. Money represents how much real wealth you could have if you spent it. It's deferred consumption. There is no real wealth in something that didn't happen.
    - Cipater 5y ago
      Money is a store of value. Money is a measure of wealth.
- mcherm 5y ago
  It does not. Bitcoin (and it's successors) are a decentralized solution to the double spend problem. If "decentralized" is a requirement, then a cryptocurrency can be an excellent solution. Other known solutions (like "just trust everyone") don't scale well (an understatement!). If "decentralized" is not a requirement then there are other solutions to the double-spend problem.
  - mdoms 5y ago
    So it's your contention that the current traditional financial system "don't scale well (an understatement!)"?
- snarf21 5y ago
  Because the trust is distributed. One single person can't cause the entire system to break down or steal all them money. With blockchain coins, the incentives are aligned. The weird thing that happened with BTC that I don't think Satoshi could forsee is that the more valuable it got, the less usable it got. This is one of the unsolved issues: how to create decentralized consensus very very quickly that can't also be attacked via a PoW that is too easy.
  - ironSkillet 5y ago
    Do you know of any lines of research that are out to prove mathematically that it is impossible to have these desirable characteristics simultaneously? I.e. any improvement in consensus efficiency necessarily gives up some amount of decentralization, appropriately defined.
    - ItsMonkk 5y ago
      This is the sort of thing that really needs to be questioned. 1. How much security do we need for this transaction? 2. For this block of transactions? 3. Can we have varying levels of security? What we are ultimately trying to know is, is it profitable to double spend? If the answer is yes, then no one should be transacting, and anything they do transact is a gamble. If the answer is no, there's much less risk, and the farther the answer is no, the safer the transaction. But we likely don't need as much security as we do now. In fact, it's very easy to handle at a user level. A. Put the transaction on the blockchain. B. Wait for additional blocks. C. If your transaction is huge, wait for more additional blocks. Wait a week, a month. This is an important transaction. When the blockchain has not reverted for such a long time, finally do the transaction in the real world. If you are transacting with yourself, you don't need any security at all. If you are transacting with a bar and the bar knows you, much like a tab, you don't need much security. If you are transacting with an anonymous person, you want high security. So security is currently handled on a blockchain by blockchain basis, where it should be a transaction by transaction basis. What if you can't wait a month? Then you should put heavy fees on that transaction and it should be put in with other higher fee transactions. 4. How does security change with transaction size? Does it make sense to have limits? 5. How does security change with total transaction block size? Does it make sense to have limits? 6. When the coin price goes up, how does that change these factors? 7. When the mining reward in coins drops, how does that affect security? 8. When Bitcoin, in 100 years, moves to 100% transaction fee system, what happens to security? What would the transactions fees look like? 9. In 50 years when the block reward has been halved 5 times, either the price of Bitcoin moves to $1 million, or the security drops. Is that okay?
- wyager 5y ago
  If you’re fine with all the downsides of trusting you bank, the bank’s central bank and currency issued, the bank’s government, etc then that obviously makes Bitcoin less appealing for you. Remaining upside includes reduced friction for things like international settlement. Don’t tell me that international settlement is easy with traditional banks - I am a financial system power user and it is not.
cryptica 5y ago
I think this is an accurate assessment but I think authentication and account management should be seen as separate from 'identity'. Authentication can be useful without being associated with a real-world identity; for example, it can be useful for spam-prevention as an alternative to signups with captcha. You could theoretically log into multiple services which you don't trust by signing messages using your private key without revealing your passphrase to any of those services.

cryptica 5y ago

Authentication using the blockchain makes a great AI-resistant alternative to signups with captcha. The same account could be used by multiple independent services and this removes the need for a signup process altogether. It also removes the need for password management since you can use a single account/passphrase to log into various services without revealing your passphrase to any of these services; you can simply sign a message to prove to any service that you control an account (and know the passphrase) without actually revealing the passphrase to any of the third party services.

UncleMeat 5y ago
> The same account could be used by multiple independent services and this removes the need for a signup process altogether. This is only marginally better than SSO with Google/Apple/Facebook/whatever. It dodges the "Google closed my account" problem, which is a meaningful improvement. But I'm not fully certain why you need the public key to be part of a blockchain for this to work. Couldn't you just publish your public key identity on any number of considerably less wasteful systems? As far as I can tell, the only advantage here is the mapping between the name "UncleMeat" and my public key would be able to be reused across services... but in so many cases I don't want my identity to be consistent across services.
- cryptica 5y ago
  - Resistant to centralized censorship (as you pointed out) - It allows you to use the same passphrase for all services without compromising security since your passphrase never needs to leave your own machine; it is never sent over the wire. You just send signatures to different services. - The cost associated with purchasing tokens needed to initialize an account on the blockchain would serve as a spam prevention mechanism; an alternative to SIM cards which most centralized services rely on today as a cost barrier to limit the creation of spam accounts. - Superior integration potential between different services/systems provided by different companies/groups since they can all refer to the exact same account on the same blockchain and provide new ways to unify data between the different services.