4 ms·

It's absolutely just a buzzword. A blockchain is just a data structure which is useful in Nakamoto Consensus, which is only for censorship resistance. Everythi

It's absolutely just a buzzword. A blockchain is just a data structure which is useful in Nakamoto Consensus, which is only for censorship resistance. Everything a cryptocurrency can do be done much more efficiently, faster, cheaper, safer, better UX, etc, except for censorship resistance. If you don't want your coin/token to be censorship resistant then you can have a system that's a million times better by just building a web app that uses regular cryptography.

stale2002 8y ago
Not really. If you build a web with a private, centrally controlled database, then you might be messing with the database behind the scenes. A Blockchain makes that database transparent, and makes any secret funny business immediately apparent.
basch 8y ago
That black or white worldview absolutely ignores a middle ground use case for a bunch of collaborative entities that _sort of_ trust each other, are ok being non-anonymous, and want to work together on a shared ledger. Ten firms working towards a common goal can do so without an insane amount of redundant double entry accounting at each firm, they can do the work once and share the results, they can audit and verify the log. And they arent all just accessing one databased stored at one firm. Im not sure people really understand the implications of a team of companies that collaborate often banding together, and eliminating the need for 3rd party audits and 3rd party fund settlement.
- sparkie 8y ago
  > Im not sure people really understand the implications of a team of companies that collaborate often banding together, and eliminating the need for 3rd party audits and 3rd party fund settlement. I think it is more that people don't understand what existing technologies there are already, but they've heard of this "blockchain" thing, and the black and white view is one where you either need a blockchain or you don't need anything. There are some other interesting technologies and research areas. One example is SMPC. In 2008 (pre-dating Bitcoin), a secure auction with sealed bids was conducted in Denmark for the purchase of beets from independent farmers[1]. This kind of research get drowned out by the blockchain hype because nobody is interested in something which doesn't print their own money. There are also more well understood distributed databases and consensus algorithms which work when you have semi-cooperative parties. I think a lot of these "blockchains" really just want a merkle-tree, and they could probably get by with a database like noms[2] When you have a bunch of cooperating parties, they either need to assume that the parties are cooperative, or you need an arbitration process. The arbitration process is what the blockchain is about (with the record keeping in the blockchain being a kind of toxic waste which is kept around to support that). What do you do when the parties are being uncooperative? If there are only a few parties, then PoW is not a viable solution because one party could outpace others if they have more computing power. If you pick something like Proof-of-stake, then several parties could collude to swindle others. If you have a centralized system, you've not gained anything. Ultimately, the companies involved are always going to take any quarrels to a legal court and have a human decide the outcome - meaning their blockchain is going to have to be "fixed" by manual intervention anyway. [1]:https://eprint.iacr.org/2008/068 https://eprint.iacr.org/2008/068 [2]: https://github.com/attic-labs/noms https://github.com/attic-labs/noms
- tylersmith 8y ago
  You can interact with people you don't trust using standard cryptography. The blockchain just facilitates censorship resistance.

stale2002 8y ago

Not really. If you build a web with a private, centrally controlled database, then you might be messing with the database behind the scenes. A Blockchain makes that database transparent, and makes any secret funny business immediately apparent.

basch 8y ago

That black or white worldview absolutely ignores a middle ground use case for a bunch of collaborative entities that _sort of_ trust each other, are ok being non-anonymous, and want to work together on a shared ledger. Ten firms working towards a common goal can do so without an insane amount of redundant double entry accounting at each firm, they can do the work once and share the results, they can audit and verify the log. And they arent all just accessing one databased stored at one firm. Im not sure people really understand the implications of a team of companies that collaborate often banding together, and eliminating the need for 3rd party audits and 3rd party fund settlement.

sparkie 8y ago
> Im not sure people really understand the implications of a team of companies that collaborate often banding together, and eliminating the need for 3rd party audits and 3rd party fund settlement. I think it is more that people don't understand what existing technologies there are already, but they've heard of this "blockchain" thing, and the black and white view is one where you either need a blockchain or you don't need anything. There are some other interesting technologies and research areas. One example is SMPC. In 2008 (pre-dating Bitcoin), a secure auction with sealed bids was conducted in Denmark for the purchase of beets from independent farmers[1]. This kind of research get drowned out by the blockchain hype because nobody is interested in something which doesn't print their own money. There are also more well understood distributed databases and consensus algorithms which work when you have semi-cooperative parties. I think a lot of these "blockchains" really just want a merkle-tree, and they could probably get by with a database like noms[2] When you have a bunch of cooperating parties, they either need to assume that the parties are cooperative, or you need an arbitration process. The arbitration process is what the blockchain is about (with the record keeping in the blockchain being a kind of toxic waste which is kept around to support that). What do you do when the parties are being uncooperative? If there are only a few parties, then PoW is not a viable solution because one party could outpace others if they have more computing power. If you pick something like Proof-of-stake, then several parties could collude to swindle others. If you have a centralized system, you've not gained anything. Ultimately, the companies involved are always going to take any quarrels to a legal court and have a human decide the outcome - meaning their blockchain is going to have to be "fixed" by manual intervention anyway. [1]:https://eprint.iacr.org/2008/068 https://eprint.iacr.org/2008/068 [2]: https://github.com/attic-labs/noms https://github.com/attic-labs/noms
tylersmith 8y ago
You can interact with people you don't trust using standard cryptography. The blockchain just facilitates censorship resistance.