"Bitcoin, Explained 72: Inscriptions"
This episode basically explain how arbitrary data is store on the blockchain throughout history:
- Before op return, there are fake utxos/transactions/bounded unspent tokens, which is on chain data pretend to be a transaction. Op return offers 80 bytes total size to store data, which is lightly more attractive than those fake transactions. What's important about this is that transactions are stored in RAM for easy access in the consensus code, so fake transactions also increase unnecessary overhead across the network.
- How inscription bypass this is by storing data in the input/witness data; This doesn't need to be in RAM once the utxo is spent. Only outputs(transactions before spent) need to be in RAM constantly. This is enabled by taproot.
- There are some controversies around this, one is that average node runner who doesn't use the inscription software(used to decode that inscription data), cannot detect what are those inscription data(the cost is too high). So they won't be able to strip that out that part, which take up extra bandwidth across the whole network. If those data got strop out, there's another problem, the mempools won't be consistent so that it's hard to verify past transactions. This splits the network.
-The controversy around taproot is that, taproot is supposed to make transactions easier to bound together and easier to process. But the current reality is that nodes are allowing all kinds of images on chain, that makes normal transaction way more expensive.
#BTC #Blockchain #Ordinals #Inscription #Taproot