Bitcoin zero-confirmation transactions are facing a crisis. Is the “replacement of fee method” really necessary?

The recent addition of Peter Todd’s Replace by Fee (a.k.a. RBF) proposal to Bitcoin Core has reignited the discussion in the community, and over the past few days, there has been an increasing amount of discussion involving both Replacement by Fee and zero zero-confirmation confirmation transactions.

This post is not intended to debate the merits of RBF itself. Personally, I don't see any advantages to it, but I'm open to the benefits it may have. I don't know that smarter people than me can debate its merits. I think Peter Todd is a very smart guy, and he's someone I'm always eager to learn from.

The purpose of writing this article is just to highlight the importance of zero confirmation transactions in the Bitcoin world. Some people in the community are too quick to dismiss it or even denigrate it.

The consequences of RBF are still controversial, but it is fair to say that RBF makes some zero-confirmation transactions more likely to be double-spent, and possibly all zero-confirmation transactions (see the RBF description at the bottom of the article for details). This means that zero-confirmation transactions are too risky to be accepted. For every Bitcoin transaction, it may be necessary to wait for a block to be confirmed.

So, is this a problem?

In my humble opinion, Bitcoin will become more dangerous and worse because of zero-confirmation transactions, and the ecosystem will become more barren. The average confirmation time for the entire industry to complete a transaction will increase due to "reasonable security", and the friction of transactions will increase. I am not sure what advantages RBF has (what are they, ask again?) that can make up for its disadvantages.

Many people in the community (including many very smart people) have long said that zero-confirmation transactions should not exist. They say that zero-confirmation transactions are always unsafe, and anyone who accepts zero-confirmation transactions is a fool, and they will eventually learn their lesson. For these people, RBF has no disadvantages because they do not think zero-confirmation transactions are necessary.

I understand this point of view. It is true that zero confirmation is risky, and it has been verified by many people and will happen in the future.

I know this because I have been a victim of it countless times.

Let's go back to the days of SatoshiDICE, which has become somewhat legendary. In the second half of 2012, SatoshiDICE was responsible for more than half of all Bitcoin transactions on the planet.

SatoshiDICE is popular for many reasons, one of which is that it provides players with a frictionless, near-instant service. There is no need to register an account or deposit funds, and players anywhere on the planet can place bets and receive their winnings (or be notified of losses).

A big reason why SatoshiDICE has been able to thrive is because it embraced zero-confirmation transactions, which kicked off the universe of Bitcoin gaming sites and helped the industry get out of the stagnation of 2012. Instead of waiting 10 minutes (or 30-60 minutes for block confirmations), players can get results instantly. Imagine if you walked into a casino in Las Vegas and had to wait 10-30 minutes every time you put a token into a slot machine to activate a game, Las Vegas would not exist.

On the surface, it doesn’t seem like a big deal, and ten minutes isn’t a lot of time. But ten minutes can be the difference between “Wow, this is awesome,” and “Okay, let’s go do something else,” especially when 10 minutes isn’t enough and it becomes 30 minutes. People’s attention spans online are even shorter. If you’re reading this article for more than just the headline or the comments, you have a much better attention span than a lot of people.

Zero confirmation was a critical part of the user experience for SatoshiDICE since its founding. The same is true for ShapeShift today, BitPay, or digital content delivery services, or tokenized access, or any of the countless blockchain business models that have yet to be born. We (ShapeShift, or SatoshiDICE before) have been constantly working to combat the risks of zero confirmation transactions, and it is a challenge that is worth our time and money. Sometimes, we also encounter the problem of double spending. We have lost a lot of money by accepting zero confirmation deposits.

But, and this is the key, we did more work and the user experience was great. And making the user experience better can take Bitcoin out of the niche and into the mainstream. You know why CoinBase and Circle became so popular? Because they understood this priority.

This is something that is not understood in many debates: while 0-confirmation transactions are indeed risky and will almost certainly result in a loss of $X, if it improves the user experience and the benefits gained from doing so outweigh the loss of $X, then it is still a positive. The risk cost of 0-confirmation must be taken into account. If your logic is, "0-confirmation is risky, therefore no one should do it," then you are wrong because you are only looking at the cost of the feature, not from a benefit perspective.

Some Bitcoiners will denigrate zero-confirmation transactions because they "carry significant risks" while they hold Bitcoin as an asset (Bitcoin is one of the riskiest financial assets out there). Owning Bitcoin as an asset is as risky a proposition as accepting zero-confirmation transactions... so do we do nothing? I advocate trying both when they are perfected.

For some use cases, the risk of zero confirmation is simply too great. The same is true for other applications. No one would recommend selling a $20,000 car with zero confirmation, but an ice cream vendor would be happy to accept the risk of a double spend to sell you an ice cream.

Clearly, as the Bitcoin economy becomes more widespread and diverse, there are some business models that will benefit from zero-confirmation transactions. For them, zero-confirmation, despite its risks, is more like a feature than a bug.

This means that if zero confirmation is effectively broken, such as through RBF schemes or any other future mechanism, Bitcoin commerce will shrink or even disappear. We can debate how big this segment will be, but you need to be careful about your assumptions. More importantly than any single commerce today, the commerce that is "unseen" in the future will disappear forever because zero confirmation is destroyed and no one will be allowed to risk using such a useful tool.

In my opinion, the current implementation of Bitcoin provides a very elegant balance between instant settlement and security. Zero confirmation should always be an option for merchants and individuals. It has risks, but in many cases, many companies are trying to manage this risk, such as BlockCypher. For those who do not want to take any risks, they can decide to wait for one or even several confirmations. SatoshiDICE can take the risk of zero confirmation without organizations infringing on their rights, and other companies can also decide to wait for block confirmations. Wouldn't it be nice to live in harmony?

However, if Bitcoin were changed in such a way that zero confirmations were completely unusable, there would no longer be a choice. We would just have to wait 10, 20, 30 minutes. This wouldn't affect some businesses, but others? They would be hampered or disrupted. And potential killer apps that are unimaginable today would never blossom. The choice between user experience and security would no longer be flexible for business: it would become rigid and immutable because of the protocol. Sometimes this is wise, but sometimes it's not.

Maybe RBF is a sensible, much-needed new feature. Maybe RBF won’t kill zero-confirmation. The point of this article is to simply convey the importance and value of zero-confirmation transactions, and that in some cases, it can be a logical choice and should not be deliberately dismissed.

Consider that if the next evolution of Bitcoin requires a reduction in usage, such as destroying zero confirmation, then other blockchain assets will fill the gap (Litecoin, or any other?). Of course, this is fine, there are many digital assets on ShapeShift. However, I want to briefly appeal to the community because no one wants to make Bitcoin worse.

If RBF does have any benefits for Bitcoin, that would be great. However, it has to be an advantage that clearly outweighs the benefits of zero confirmation.

It would be unwise to abandon zero confirmation simply because of its (RBF) advantages.

Some background on (RBF)

RBF is a modification to Bitcoin confirmation processing. Currently, unconfirmed Bitcoin transactions are stored in the memory pool (mempool). Under the rules of the RBF scheme, if the fee of the second transaction is higher, these unconfirmed transactions can be replaced by a later double-spending transaction. Simply put, before the RBF scheme, if you double-spend two transactions, miners will usually ignore the second transaction because they follow the "first seen" rule. Therefore, the first transaction seen "may be confirmed", while under the RBF rule, when a double spend occurs, the transaction with the highest fee will be confirmed. Therefore, the first transaction seen is meaningless.

An important caveat is that only transactions that follow the new rules will have the “RBF mark” added, while transactions without the RBF mark will still follow the “first sight” rules.

This means that transactions sent by wallet software that adds the RBF flag by default may never accept zero-confirmation transactions, because any subsequent double-spending transactions, if one of them includes a higher fee, will replace the original transaction, thereby deceiving the recipient.

In other words, RBF-enabled wallets always require 1 confirmation, because the risk of double spending has gone from "possible" to "almost certain." If most wallets and miners adopt this RBF policy, zero-confirmation transactions will disappear on their own.

Original: https://shapeshift.io/site/blog/2015/12/01/note-ceo-erik-voorhees-appeal-zero-conf
By Erik Voorhees
Compiled by: Satuoxi
Editor: Satuoxi
Source (translation): Babbitt Information (http://www.8btc.com/appeal-zero-conf)