double-spending problem？

[shuijian-xu]

Transactions transfer bitcoins from one party to another and are unlocked, or authorized, by signatures. This ensures that the sender authorized the transaction, but what if the sender sends the same coins to multiple people? The owner of a lockbox may try to spend the same output twice. This is called the double-spending problem.

[shuijian-xu]

This is where a major innovation of Bitcoin comes in, with blocks. Think of blocks as a way to order transactions. If we order transactions, a double-spend can be prevented by making any later, conflicting transaction invalid. This is the equivalent to accepting the earlier transaction as the valid one.

Implementing this rule would be easy (earliest transaction is valid, subsequent transactions that conflict are invalid) if we could order transactions one at a time. Unfortunately, that would require nodes on the network to agree on which transaction is supposed to be next and would cause a lot of transmission overhead in coming to consensus. We could also order large batches of transactions, maybe once per day, but that wouldn’t be very practical as transactions would settle only once per day and not have finality before then.

[shuijian-xu]

Bitcoin finds a middle ground between these extremes by settling every 10 minutes in batches of transactions. These batches of transactions are what we call blocks. In this chapter we’ll review how to parse blocks and how to check the proof-of-work. We’ll start with a special transaction called the coinbase transaction, which is the first transaction of every block.