I think it would be possible to create a legitimate seeming transaction on the Bitcoin testnet that the Bitcoin mainnet would accept.
I've read a bit about the ETC replay attacks, and it seems like it would be impossible, since there was never any hard fork to create testnet.
https://medium.com/@timonrapp/how-to-deal-with-the-ethereum-replay-attack-3fd44074a6d8
But I'm not sure about this. Has anyone successfully done it?
Do you know a good reason why it 100% would not work?
The answer is no, you can't, because the two chains don't share a common history.
Internally, transactions don't spend "from" an address. Instead, they refer to a specific output created by an earlier transaction which they consume.
In order for a replay attack to work, that same transaction whose outputs are being spent must exist on both chains. This is only possible when the two chains share a common history, and the earlier transaction either occurred before the fork (so it exists on both sides), or it itself was already a replay.
As soon as a transaction's history is not identical - meaning it consists of the same transactions leading up all the same to the same coinbase outputs that created the coins - they cannot be replayed, as the txids of the inputs will not exist in both chains.
I think it would be possible to create a legitimate seeming transaction on the Bitcoin testnet that the Bitcoin mainnet would accept.
no
