15 ms·
"transaction reversibility is not about the ledger, but rather about the transaction rules that a currency uses. A reversible currency requires that someone ano
by JackC 4y ago
"transaction reversibility is not about the ledger, but rather about the transaction rules that a currency uses. A reversible currency requires that someone anoint this trusted party (or trusted parties) and that they use their powers to freeze/burn/transact currency in ways that are at odds with the recorded owners’ intentions. And indeed, this is a capability that many tokens now possess"
I think this is arguing that reversibility is not antithetical to permissionless blockchains, because reversibility can be implemented on top of permissionless blockchains.
But that doesn't answer the critique -- the critique is that most real world systems do need trusted parties, and if you build a system of trusted parties on top of a permissionless blockchain, then you could have saved a lot of complexity, risk, and proof-of-economic-waste burn by building on top of a permissioned distributed ledger instead. Reversibility is one example of a design requirement that undoes the claimed advantages of permissionless blockchains that are meant to justify their inherent downsides.
- randomran01234 4y agoAnd it is fine to build systems with trusted parties on top of a permissionless network. The article cites USDC which does this, it is one of many applications users can choose to transact with. Attempting to bake reversibility into the protocol would lead to a tightly permissioned system with only a few trusted authorities. A loose analogy might be the internet, which aims to be a decentralized global protocol that we can then build centralized systems on top of.
- adra 4y agoThe internet is built on the 'A' server which is a tightly guarded centralized US tool. Commerce lives because the absolute power brokers of the system have chosen to not exercise strong enforcement. The internet technologies are certainly built in a way that could function in completely autonomous blobs all over the world, but that's certainly not the way it functions today.
- lbriner 4y agoI'm not an expert but I think this is what the original letter authors objected to. If you build these mechanisms on top of a blockchain which is inherently baked in stone, then there is no point in having the blockchain be so strict. In other words, if you allow an organisation to change things like refund you, then they might as well just manage their own leger like they do at the minute.
- randomran01234 4y agoThe goal of modern crypto like Ethereum is to provide a secure, predictable, open source, and decentralized network that many higher-level applications can flourish on top of and co-exist within. USDC exists with reversibility but not all users in the network are forced to use it. There are other protocols like DAI that have different features and considerations. And in many cases these are open source protocols that can be forked as desired. Probably the closest thing we have like this is the web, which is more or less a decentralized protocol, atop which we have built a lot of centralized platforms.
- rvz 4y agoThis probably the likely case of having co-existing protocols like Ethereum, USDC and some Layer 2s. They don't seem to be ready yet but they look somewhat promising.
- pcthrowaway 4y agoThere's a benefit to building it on the blockchain. I can look at the USDC contracts and see that privileged users have the ability to freeze the USDC in my account. I can look at the DAI contracts and see that they do not. Good luck getting visibility into the back-end processes of web2 applications. Even if they publish the source on github, there's no way to verify what they're running
- dmitriid 4y ago> Good luck getting visibility into the back-end processes of web2 applications "Smart" "contract" are as invisible to the average user as the backends of web2 applications (or any other applications for that matter). The authors of these "contracts" routinely create buggy contracts because the code is complex [1] But sure. You can definitely look at impenetrable code written in an esoteric language for an equally esoteric VM and see exactly what it does. [1] just an example, https://web3isgoinggreat.com/?id=akudreams-earns-34-million-team-will-never-be-able-to-withdraw https://web3isgoinggreat.com/?id=akudreams-earns-34-million-...
- NoGravitas 4y agoIf you build a system with trusted parties, you no longer need to build it on top of a permissionless network. If permissionless networks were zero-cost and had no downsides, it wouldn't matter. But in fact, they are incredibly expensive and complex compared to conventional systems (like relational databases), and have serious downsides. Once you need a trusted party, there's no way to justify the permissionless network.
- tornato7 4y agoWell, I can think of a few reasons: facilitating transactions with other parties on the permissionless network, promoting public verifiability of your system, utilizing the tools and utilities available for public blockchains, requiring your data to be available even after you stop paying your hosting provider, avoiding the hassle of building your own payment on-ramps, etc.
- dcolkitt 4y agoMany real world systems are "nearly trustless". Of course you still need some sort of court system if someone decides to break bad. But in the 99%+ of times you're in the happy path, economic transactions occur based on autonomous rules encoded in software. The analogy I like to use is what's harder to buy a $1 million house or $1 million of Microsoft stock? The former process takes weeks, and dozens of man hours from lawyers, realtors, escrow agents, bankers, county property registrars, etc. Whereas a buy order to execute Microsoft shares happens in microseconds. That's because we've built a highly streamlined autonomous system for executing stock transactions. Yes of course, there's a trusted layer built on courts and fiat law at the bottom. And occasionally for a corner case, real lawyers have to get involved. But the vast majority of the time, this layer is completely abstracted away and trading stocks are just bits flipped on a computer. When somebody says something like "let's put houses on the blockchain", the idea isn't to get rid of the fiat law system completely. At the end of the day, you still need courts and sheriffs to enforce property rights. But the idea is that we can wrap the house in an LLC, whose bylaws state that it's governed by on-chain contracts. (Which Delaware Chancery courts will absolutely recognize and enforce.) Now instead of directly transacting at the very low-tech county property registrar we can transact using an on-chain NFT that grants the equivalent of ownership rights to the house. (This is very analogous to how Cede & Co technically owns almost all the stock shares in America, and holds them under your name for you.) That NFT is way more powerful and efficient to transact with than the county property record. It lives on a credibly neutral level that already has billions in native capital and liquidity and exposes a fully Turing complete smart contract system. You can sell your house atomically and instantly to anyone in the world with no lawyers or escrow agents. You can pledge it as collateral and borrow against it. You can fractionalize it and sell a portion. Or roll it into a larger portfolio and slice into tranches. All with no more than few dozen lines of code. Without blockchains all of those operations would take hundreds of pages of legal documents and weeks of back and forth.
- what-the-grump 4y agoReduction of complexity somehow yields mobility and fractionalization of a complex physical asset? Why is this a good thing when you skip all the protection, nearly 500 years of property law and understanding all so you can digitize something into meaningless bits? Replace NFT with deed, and what you think we can’t execute the transfer quick enough? No we can but we don’t because, we want to title search, violation search, inspect, appraise, survey the freaking thing you are about to blow your life’s savings on. There is a reason why people continue to lose millions and billions on NFTs and crypto and why systems to slow and prevent rampant fraud exist. There is no real technical problem with a real estate transaction that has not been solved. I closed on my house remotely, in 6 weeks, never having met anyone except my broker. 6 weeks were spent doing due diligence inspections. I have an NFT of the Brooklyn bridge to sell you, cheap, 200 million dollars.
- roca 4y agoThis is such an obviously valid objection that I'm disturbed the author didn't address it inline.
- rkagerer 4y agoThe missing piece crypto lacks is an opt-in decentralized justice system for resolving disputes.
- danuker 4y agoCheck out Bisq. https://bisq.wiki/Frequently_asked_questions#Dispute_resolution https://bisq.wiki/Frequently_asked_questions#Dispute_resolut...
- astoor 4y agoTransaction irreversability is the whole point of the "peer to peer electronic cash system": "Commerce on the Internet has come to rely almost exclusively on financial institutions serving as trusted third parties to process electronic payments... cutting off the possibility for small casual transactions ... What is needed is an electronic payment system based on cryptographic proof instead of trust, allowing any two willing parties to transact directly with each other without the need for a trusted third party. Transactions that are computationally impractical to reverse would protect sellers from fraud" from the opening lines of the Bitcoin white paper. So Bitcoin was designed for micropayments, and irreversability is a feature to reduce the friction. That is fine because it was intended for "small casual transactions" which very few people are going to invest the time and effort into disputing. It even looked like it might be successful as a micropayment system at first, given small Bitcoin transactions were initially processed without any transaction fees. The problem is that it has clearly failed as a "peer to peer electronic cash system". It is now used primarily for large transactions, which you absolutely do need consumer protections for if you are a legitimate user (indeed the fact that there aren't consumer protections has made the space so popular with fraudsters, scammers etc.). And as others have commented, the newer cryptocurrencies which attempt to offer such protections end up being worse in every conceivable way from the traditional solutions. Leading back to the original article - is there any legitimate point to cryptocurrencies nowadays?
- DebtDeflation 4y ago>It is now used primarily for large transactions No it's not. It's now used primarily as a vehicle for speculation.
- lumannnn 4y agoYou probably should look at different numbers than just the dollar price. Number of nodes is rising, number of wallets is rising, number of hash-power is rising. Also, please look into the "Lightning Network". It's the second layer on top of Bitcoin and that's where the whole ecosystem scales (in terms of numbers of transactions per second). Cheap, scalable and fast transactions. Adoption is happening in many places.
- cy_overlord 4y agoThe argument made was that it's not about the technology itself but the transaction rules. If money are stolen from banks in a specific way (withdrawn into prepaid cards), banks will not and cannot reverse these transactions. Banks do have some rules like chargeback however to reverse legitimate transactions or delayed transactions. The same rules can be converted into Smart Contracts which governs the transaction of a certain currency. So instead of reversing, we could make all NFTs locked from trading for 3 days after transaction which anyone could appeal for a dispute and buyback the NFT for the same price anytime. The challenge here is, can we actually come up a set of robust rules that is not too rigid yet not abusable. It's not easy, it won't be right off the bat and it certainly require a lot of critical, innovative thinking. And maybe the solution won't have to be complex at all.
- GTP 4y agoThe problem that I see here, is that with the "code is law" approach if there's a bug in the smart contract that is supposed to protect you then you're screwed without recourse. In a traditional system, if there is a bug in your bank's software that makes you loose money, you can ask the bank to give you the money back. If they refuse, you can take them to court. If instead the smart contract has a bug, there's no one you can appeal to to get your money back. So I think that having humans-in-the-loop is still important, at least for big transactions. As others pointed out, this problems are much smaller for small transaction so maybe there could be some use there. But I have to think better about this use case.
- cy_overlord 4y ago"you can ask the bank to give you the money back" "you can take them to court" Which like the article argues, is not something that can be solved by any ledger, distributed or not. Even if banks reverses records, they don't change the data which is already on the database/record, instead they just issue a new transaction from the backend or withdrawing it from a legit reserve that they own. In any case, no reversing was done. What if instead of taking the bank to the court, we have something in place for it that runs based on certain "if statements" or "switch case". Of course, is not as easy as it sounds, we need conditions that are robust enough and not vulnerable to abusers. It's a very complicated problem to solve.
- once_inc 4y agoBitcoin is a base-layer solution. Do you not want to transact on the base layer because of lack of reversibility or fees? Move to a higher layer solution that sacrifices some characteristic of Bitcoin to enable something new. Want to do lots of (micro)transactions? Use LN where this is made possible by sacrificing the ability to be secure offline. Do you want to use DeFi-like solutions with smart contracts? Use RSK. I don't see what the problem is with moving things up one layer if you retain the option to delve back into the base layer.
- ninkendo 4y agoThe base layer is melting the planet, for one thing.
- once_inc 4y agoBitcoin mining isn't melting the planet. It's not even using more energy than Christmas lights or wash dryers. The comparisons made with [insert random country here] is dumb, because any worldwide phenomenon has a large chance to use more energy than [insert same random country here]. Even if the base layer of Bitcoin is using energy, one can argue that the functionality it offers (trustless, borderless, permissionless, programmable, digitally-native money) is worth the energy cost.
- dmitriid 4y ago> because any worldwide phenomenon has a large chance to use more energy than It's not a worldwide phenomenon. There are orders of magnitude more Christmas lights and wash dryers than mining servers. And yet...
- ninkendo 4y agoBut we’ve already established that we can do all that without the energy costs if we end up needing a trusted authority anyway. There’s zero need for proof of work if you’re going to trust a central authority to validate the transaction history. What’s the point of such an extremely costly base layer?