You are right. Replay attacks have to be started by an attacker. Those won't happen "automatically".
Regarding your second point: If you first (after the fork) send your bitcoins to a new address owned by you, the transaction still could theoretically be replayed.
But an attacker just had no advantage in replaying this transaction since it would sent your bch to an address under your control.
Still this can be performed by an attacker to spread fear around. Trolls are everywhere.
If you want to get sure none of your transaction gets replayed you have to send your btc and your bch to a new (not equal) address.
From this point your transactions can't be replayed since there are no inputs on the (respectively) other chain to spend.
Topic: A question about replay attacks. (Read 259 times)
No there is no need for an attacker to have your coins lost in a replay attack. A replay attack is basically when the same transaction is repeated over the other chain. This might happen due to low security and lack of replay protection. The hacker might take advantage of certain situations if we are not careful form our side. Consider not doing any transaction right after the fork and you will be safe. Wait at least 1 or 2 days after the fork for the dust to settle down. Import your private keys in a Bitcoin client which guarantees replay protection and then make any transactions.
Meaning that the unwanted transaction does not happen automatically
Right. Publishing a transaction to the network of one fork will not automatically publish to the network of the other fork. However, a replay on the other network doesn't need an 'attacker' in the sense of someone intentionally being malicious. It possible for a misconfigured node to accidentally publish it to the other network.Meaning that the unwanted transaction does not happen automatically
Right. Publishing a transaction to the network of one fork will not automatically publish to the network of the other fork. However, a replay on the other network doesn't need an 'attacker' in the sense of someone intentionally being malicious. It possible for a misconfigured node to accidentally publish it to the other network. 
I joined in on bitcoin 1 month ago so forgive me if my terminology is incorrect or if I am not clear enough.
My question is, are replay attacks initiated my an "attacker"? Meaning that the unwanted transaction does not happen automatically and accidentally, it is manually replicated on the unintended chain however way it is done, which I am clueless of.
If it is done by an attacker, let's say I had bitcoin on a wallet (1) before a fork happened, and after it happened, I transferred my bitcoin to a new wallet (2), this transaction will not be replayed as the receiving address is also mine. Then I will be able to safely claim or spend my forked coin on wallet 1 because it does not have any bitcoin in it, right?
Thanks in advance.
My question is, are replay attacks initiated my an "attacker"? Meaning that the unwanted transaction does not happen automatically and accidentally, it is manually replicated on the unintended chain however way it is done, which I am clueless of.
If it is done by an attacker, let's say I had bitcoin on a wallet (1) before a fork happened, and after it happened, I transferred my bitcoin to a new wallet (2), this transaction will not be replayed as the receiving address is also mine. Then I will be able to safely claim or spend my forked coin on wallet 1 because it does not have any bitcoin in it, right?
Thanks in advance.
Jump to: