Did Bitcoin hard fork in 2010?

Recently, due to the controversy over Ethereum's hard and soft forks, many people are discussing the early history of Bitcoin. Those who support Ethereum's hard fork claim that Bitcoin had a hard fork in 2010. This is either a half-understanding or intentional misleading. I have never invested in Ethereum, but I must seriously correct the historical issues of Bitcoin.

What are hard forks and soft forks?

To understand what happened in 2010, we must first understand the meaning of hard forks and soft forks. Both hard forks and soft forks refer to changes in the rules of transaction validity. A hard fork is when an invalid transaction becomes valid, while a soft fork is when a valid transaction becomes invalid. This definition has always been clear and unambiguous.

What happened in 2010?

In 2010, several loopholes in Bitcoin transaction processing were discovered several times.

The first time was in July. There is an OP_RETURN function in the Bitcoin programming language, which means that all checks can be skipped to make the transaction valid immediately. Therefore, anyone can use OP_RETURN to skip all checks and spend everyone's Bitcoin at will. Satoshi Nakamoto immediately corrected it and made all transactions with OP_RETURN invalid, which is what we see today.
In the same incident, it was discovered that some program functions such as OP_LSHIFT would cause some nodes to crash, so Satoshi Nakamoto used the same method to invalidate these functions.
The above attacks only occurred on the test network, so no real losses were caused.

The second time was in August. Bitcoin transactions have a rule that the total amount of output cannot be more than the total amount of input, otherwise Bitcoin can be issued at will. However, someone discovered a design loophole at the time. It turned out that the transaction output could be negative, so as long as one of the two outputs was negative, the amount of the other output could be higher than the total amount of input. When this transaction was confirmed on the main network, Satoshi Nakamoto issued an urgent correction to prohibit negative output values. After the miners upgraded, they abandoned the block with the negative output transaction, so no one will see that transaction today.

Is the above a hard fork or a soft fork?

In July, transactions with OP_RETURN, OP_LSHIFT, etc. were changed from valid to invalid; and in August, negative outputs that were originally valid were changed to invalid. In the whole process, no invalid transactions were changed to valid, so they were all soft forks.

I want to reiterate: Bitcoin only had a soft fork in 2010, not a hard fork. What the future of Ethereum is is not what this article is about; but when discussing it, everyone must recognize the historical facts and not be misled, nor mislead others.