4 ms·

My point is that lookups and simple `if then else` comparisons can be used to do verification too with similar performance characteristics and a lot simpler to

My point is that lookups and simple `if then else` comparisons can be used to do verification too with similar performance characteristics and a lot simpler to implement. Is it related to zero-knowledge proofs? The industry makes it seem like Merkle trees are required for all blockchains which is definitely not the case since most blockchains are public and each full node has a full record of the history.

saagarjha 7y ago
I mean, each node doesn't have a literal full record of the history embedded in it, right? There's just a back pointer to the previous node. And the Merkle Tree part verifies that the thing the node points back to is legit and not just a fake block you pulled out of thin air.
- cryptica 7y ago
  I was talking about nodes as independent machines which have the full history (full nodes). Maybe you were thinking of nodes in a mathematical sense like blocks in a blockchain. My point stands that if each block ID is computed from: 1. A hash of all the transactions inside the block. AND 2. The hash from the previous block. Then you can verify that the block is legit without a merkle tree. Tampering or substitution of any block would result in the hash not being consistent with the neighboring blocks.
  - saagarjha 7y ago
    I mean, isn't that literally a Merkle Tree?
  - krcz 7y ago
    But in order to prove that to someone that doesn't have full history, they would have to synchronize it, by downloading all the data and verifying all the transitions. With Merkle tree (identified with its root hash, just as blockchain is identified by hash of its head) it is possible to verify there is a value inside without full synchronization.
  - pests 7y ago
    It's not about Bitcoin or blockchains and nothing about zero knowledge proofs or anything. You did just describe a merkle tree though. Here is another example. BitTorrent uses merkle trees for torrent and piece identification. You request peer list from a tracker via the torrents merkle tree root hash. These are visible in torrent files and sometimes magnet links. That root hash corresponds to a certain tree of piece hashes whose data that when reassembled will recreate the torrent. The tree of piece hashes is included with the torrent file. You can now ask peers for those pieces by their hash and know the peer sent you the correct data because it hashes correctly. Furthermore you can know you reassembled the original torrent because you can verify all the piece hashes hash back up to the merkle tree root this verifying the torrent was downloaded correctly.

saagarjha 7y ago

I mean, each node doesn't have a literal full record of the history embedded in it, right? There's just a back pointer to the previous node. And the Merkle Tree part verifies that the thing the node points back to is legit and not just a fake block you pulled out of thin air.

cryptica 7y ago
I was talking about nodes as independent machines which have the full history (full nodes). Maybe you were thinking of nodes in a mathematical sense like blocks in a blockchain. My point stands that if each block ID is computed from: 1. A hash of all the transactions inside the block. AND 2. The hash from the previous block. Then you can verify that the block is legit without a merkle tree. Tampering or substitution of any block would result in the hash not being consistent with the neighboring blocks.
- saagarjha 7y ago
  I mean, isn't that literally a Merkle Tree?
- krcz 7y ago
  But in order to prove that to someone that doesn't have full history, they would have to synchronize it, by downloading all the data and verifying all the transitions. With Merkle tree (identified with its root hash, just as blockchain is identified by hash of its head) it is possible to verify there is a value inside without full synchronization.
- pests 7y ago
  It's not about Bitcoin or blockchains and nothing about zero knowledge proofs or anything. You did just describe a merkle tree though. Here is another example. BitTorrent uses merkle trees for torrent and piece identification. You request peer list from a tracker via the torrents merkle tree root hash. These are visible in torrent files and sometimes magnet links. That root hash corresponds to a certain tree of piece hashes whose data that when reassembled will recreate the torrent. The tree of piece hashes is included with the torrent file. You can now ask peers for those pieces by their hash and know the peer sent you the correct data because it hashes correctly. Furthermore you can know you reassembled the original torrent because you can verify all the piece hashes hash back up to the merkle tree root this verifying the torrent was downloaded correctly.