5 ms·
I wondered about that, too, and thought of a difference: if you sign it over to someone else with PGP, the first owner still has a file that was provably sold t
by dangoor 5y ago
I wondered about that, too, and thought of a difference: if you sign it over to someone else with PGP, the first owner still has a file that was provably sold to them. They could fraudulently sell it again. A public blockchain would contain the whole chain of ownership, so you could tell who the current owner is.
- stagas 5y agoCan't they have a public personal ledger somewhere for people to audit? Saying I sold X to Y. Like a tweet in a personal account. Why is the common blockchain necessary? Seems like centralization of information, is this what Web3 is about? Power to the gatekeepers?
- qqii 5y agoAs long as twitter can act as a trusted third party, tweets are a completely valid ledger as shown by Keybase. Blockchains simply reduce the power and trust required in the third party by offering a different set of tradeoffs.
- stagas 5y agoYou don't really need Twitter, you need a platform and require the people that use the platform to bring in their ledgers so they are auditable by the public. I can view someone's work and examine if they're trustworthy or not. Similar to how people are signing their signatures, they could be just bringing together their ledger, which could be a bloom filter so it's just a single line of hash that I can verify I still own the thing they selled, you only need a system that can compare these and see if anything has been double spent. The point is, you don't need a running bitcoin-style blockchain for this. You can achieve the same validity with different means.
- qqii 5y agoUsing a bloom filter as a distributed hash means you're edging quite close to a blockchain, the only question your schema now has is consensus - how is work examined for trustworthiness? What prevents network spam? I haven't puzzled it though but intuitively it feels like you wouldn't be able to achieve the "same validity" (aka you'd have a weaker set of guarentees) without the liveness of a blockchain like system. Whether you agree with what threshold of validity is acceptable is a different matter.
- DarylZero 5y agoYeah except you could sell the NFT on a different blockchain. As far as NFTs are concerned, the existence of multiple incompatible blockchains is indistinguishable from what would be considered a network partition within a single blockchain for cryptocurrency. @stagas: > Can't they have a public personal ledger somewhere for people to audit? Then you need to coordinate on a common authority to rely upon for the timestamping. Blockchain's whole purpose is to eliminate this dependency.
- stagas 5y agoThe signed message can have a timestamp on it. The only authority necessary is the signature of the author or the previous owner.
- DarylZero 5y agoYou need an authority to enforce that the author isn't signing a timestamp that does not match the actual clock time.
- stagas 5y agoStill not convinced. I can audit the message so the only way to game this is we both have to be playing a game. Against who? A third person who we will try to sell afterwards? Same thing can happen in the blockchain as well, no-one can audit what happens off the chain which could be people coordinating.
- DarylZero 5y agoI don't know what you mean by "audit the message." Without third-party timestamping or blockchain, the same token can be sold to multiple buyers, and nothing can be used to determine which buyer is the first ("real") buyer. This is the whole reason blockchain was invented. It wasn't possible to do it at all before that.
- qqii 5y agoThis is true, except there would at least be a record on some blockchain where the artwork was first proven to exist via a hash. This is at least stronger proof against the artist and for the buyer. As it stands this isn't very valuable yet but in my mind an important milestone.