Aaron van Wirdum Live from Utrecht, this is Bitcoin Explained. Hello. Hey, Sjors.
Transcript
Sjors Provoost What's up?
Aaron van Wirdum Our favorite guest is back, Ruben.
Ruben Somsen Hello, I'm back.
Aaron van Wirdum Hey, Ruben. How are you?
Ruben Somsen Good to be back.
Aaron van Wirdum How have you been?
Ruben Somsen I'm doing great.
Aaron van Wirdum Thanks for being back. So you're our sidechain expert, kind of?
Ruben Somsen Yeah.
Aaron van Wirdum How do you describe yourself?
Ruben Somsen I change it every day, so there's no... Yeah, I guess sidechain expert is a good way of putting it.
Aaron van Wirdum Layer 2 expert, maybe?
Ruben Somsen Layer 2, yeah. Definitely, yeah, that all fits.
Sjors Provoost Complicated other things expert.
Ruben Somsen I think someone described me as a sidechain architect, I think. That was kind of funny. Architect, I like it.
Aaron van Wirdum So it's good you're back today because we're going to discuss something layer two-ish, namely federated eCash. First of all, why are we discussing federated eCash today, Sjors?
Sjors Provoost Well, because there was a blog post in November 2021 about it. And then somebody randomly brought up eCash in another chat that I'm in last week. So I think maybe we should cover that.
Aaron van Wirdum Wait, is the blog post from a year ago? Is that what you said?
Sjors Provoost November. Just a few months.
Aaron van Wirdum Oh, a few months ago.
Ruben Somsen And it's a post by Blockstream specifically.
Aaron van Wirdum Yeah, so Blockstream is going to fund a federated eCash project. Is that right?
Sjors Provoost Yes. I think that it's going to sponsor the developer who works on that project.
Ruben Somsen Yeah, his name is Eric. And he's basically researching the project or researching eCash. And Blockstream is funding it.
Aaron van Wirdum Right. So I don't know if you know this, Sjors, but I actually, we discussed this on the show before, or at least I did kinda on the episode, the Bitcoin Beach episode I did with Nicolas Burtey. They have this shared custody solution for their Lightning wallet. And then during the podcast, I sort of mentioned, you know what would be great for this? If you use Chaumian eCash for this system to add privacy. So I'm very happy to learn that someone is actually building this. That's excellent.
Sjors Provoost So really it's your idea.
Aaron van Wirdum I think I came up with it basically.
Ruben Somsen Clearly.
Aaron van Wirdum That's a joke in case people are not sure.
Ruben Somsen [unclear].
Sjors Provoost [unclear] get to explaining it, but it does seem like it actually improves upon... The thing that we're about to discuss is a useful improvement upon custodial wallets.
Ruben Somsen Yeah.
Aaron van Wirdum Yeah, first, last intro question before we get into it. So Eric was already building this for a while before Blockstream started to throw money at it or?
Ruben Somsen I'm not sure how it got started. I think, I guess he was interested in the topic and I guess he got in touch with Blockstream and my guess would be that he was already like interested in working on it. And I know at least a lot of the Blockstream people have always been very pro eCash. So it's not very surprising to me that they are interested in funding something like this. So I think they just got together and the interests aligned and there you go. I think that's what happened.
Aaron van Wirdum Yeah, eCash definitely seems like a good solution to be built on top of Bitcoin to me. So let's get into eCash itself. What is eCash? It's yours.
Sjors Provoost In a nutshell, we have to go back in time to, I don't know, 1981 or something around that time when David Chaum published a paper. He's a cryptographer, which I think of something like blinded signatures in blah, blah, blah, electronic cash, blah, blah, blah. In a nutshell, what the system that he proposed and eventually built with DigiCash, what it did is the bank or a bank issues IOU tokens. And those tokens, you could see them as a piece of text. And we'll explain how those tokens are made. And you can actually send those tokens to other people on the internet. And they can then redeem those tokens if they want to.
Aaron van Wirdum Yeah, you speak about tokens, but the point here is that eCash was essentially a technology layer to be used by banks. So the actual money is still dollars or euros or whatever the bank. you know, whatever the bank accounts are using. And then they issue tokens that basically represent a dollar and they can be exchanged at the bank for a dollar, right?
Sjors Provoost That's right.
Ruben Somsen And I think, you know, the goal was sort of, I think also to think of it as regular cash, right? Where just a digital version of having a dollar bill in your hands and giving it to somebody else, roughly.
Sjors Provoost And with the main advantage of it being somewhat anonymous. So the bank would give you... digital dollar bills in a way that they will not know what actually is happening to these bills. So if the bank gives me a dollar bill or this digital dollar bill and then I give it to Ruben, then the bank does not actually know that I gave it to Ruben. They'll know that they gave me a dollar and they'll know that Ruben redeemed a dollar with them. But assuming there's more people using the system, they will not be able to tell which dollar is which.
Aaron van Wirdum Yeah, they won't be able to tell who paid who essentially.
Sjors Provoost And now you might ask yourself, oh my God, they actually invent Bitcoin that long ago? And the answer is no, because they have this double spend problem. And the way they deal with the double spend problem is that when you receive coins, you have to actually tell the bank, hey, I just received this coin. Please make sure nobody else receives this coin ever. So that's the centralized party has to make sure that coins can only be used once. Because I think it's easiest to think about these tokens as just a piece of text. And you can copy a piece of text around, so there's nothing stopping me from giving the same piece of text to Ruben and to give it to Aaron. But as long as both of them immediately check with the bank, hey, I just received this token, the bank will make sure that the second person will get a warning, like saying, hey, you did not actually get any money, you should not offer this service. So that's how they solved the double spend problem. And now the question is...
Aaron van Wirdum What you're saying is it's still required... a central party to sort of manage everything right?
Sjors Provoost Yeah, and to synchronously manage it, right? they have to be online all the time otherwise the whole payment system grinds to a halt right?
Aaron van Wirdum But to your point I would say well, I wouldn't just say that I think that's a fact this e-cash was the first digital currency was sort of the predecessor the first predecessor to Bitcoin.
Sjors Provoost The first digital cash right so so as distinct from a digital bank account which a digital bank account, it just means that the bank knows exactly which dollar goes to which other person.
Aaron van Wirdum Yeah, my point being, if you sort of backtrace where Bitcoin came from, then I think it's fair to say eCash is sort of the first attempt of doing something like that.
Sjors Provoost Yeah, for sure.
Aaron van Wirdum So you mentioned that it offers privacy, and I want to get into that real quick. So how does this actually work?
Sjors Provoost Yeah, that's amazing, right? Because you would imagine that if you just have a piece of text, then the bank... writes the piece of text, gives it to the first user, and then they give it to the other user. So the bank knows exactly what piece of text they wrote.
Aaron van Wirdum Let's say throughout this conversation, for clarity, I'm the bank. All right?
Sjors Provoost Oh, you're the bank.
Aaron van Wirdum I'm the bank. So I'm issuing something to you, Sjors. I'm issuing a note to you. Sjors, you pay Ruben. Ruben pays me. And then I can usually, if we don't take the privacy, anonymity measures, then I can see, hey, this is the note that I gave to Sjors. So therefore, Sjors must have paid Ruben and there's no privacy [unclear].
Sjors Provoost Because there might be a serial number on it.
Aaron van Wirdum So this needs to be solved somehow. So how was that solved?
Sjors Provoost Ta-da! Yeah.
Ruben Somsen Okay, so basically how that's solved is by blind signatures. So what happens is the first person comes and gets a note from the bank. So in this case, Aaron, you give a note to Sjors. But the trick is that it's actually a number that Sjors picked, and he blinds that number, so you can't actually see it. And then you put your signature on the blinded value. And then Sjors, he can go and unblind that value after you signed it. And so now Sjors has his number that you signed, but you never saw because the only thing you saw was the blinded number. And then later, Sjors gives the number to me. And now both me and Sjors have the number. So that's not enough. If I just received a number from Sjors, well, now it's kind of both our coin. So I go and I show a number to you, Aaron. And now you see the number and you see that you signed it, but you don't recognize it because the previous one was blinded. And after you receive it, you either credit me a dollar or now we do the same thing. We create a new number. So I create a number. I blind it. I give it to you. You sign it. I unblind it. And now I have a new token. That's basically how it works.
Aaron van Wirdum Yeah, and you're using the word blinding here but that's basically just doing some sort of mathematical formula with a secret number right?
Ruben Somsen Yeah, you can think of it as, uh, adding two numbers together, two big numbers, uh, and then after you sign it, somehow one of the two numbers can be subtracted again That's not literally how it works, but…
Sjors Provoost Yeah, or a physical analogy that's being used is this carbon copy paper. So I don't know if people still remember what paper forms are.
Aaron van Wirdum You're old, Sjors.
Sjors Provoost But well, if you drive a car, then you might still encounter these pieces of paper. So if you, in Europe anyway, if you crash your car into somebody else or vice versa, you have to fill in an insurance form and you draw a picture of the situation on the form. And then there's a piece of paper below it, which is called the carbon copy. And that's where CC in email comes from, which will have the same text on it because your pen physically pushes through the paper. It's crazy. But the analogy here would be you could put a piece of carbon, this carbon paper, inside an envelope, and then you would write on the envelope, on the outside of the envelope, your signature. And that means that whatever was on the inside of the envelope now has your signature on it. So you kind of make a carte blanche signature. Of course, it wouldn't be carte blanche. There is a way to guarantee that you're not signing for $100 when you're only getting one. But that's the idea. The bank knows they signed something that is a bank note. And you can prove to other people that the bank showed it. But the bank does not know which note they signed because they can't see inside the envelope. And unlike a real envelope, they can't actually open it.
Ruben Somsen And so maybe one good thing to add here is that the anonymity set is actually depending on the value. So if you issue a $1 note, then everybody else who has a $1 note is within the same anonymity set. And if there's a $2 note, then you have yet another anonymity set. And because the bank cannot really distinguish, you know, it signs something and it can't really distinguish who it came from. The way you distinguish the values, well, there are multiple ways of doing that, but one way of doing that is to have the bank just put a different signature on a different value note.
Sjors Provoost But now you're moving beyond what David Chaum did, I think.
Ruben Somsen That is possible, yeah.
Sjors Provoost We're not actually entirely sure, but as far as I understand it, the DigiCash system had coins of one fixed value and they were all the same value.
Ruben Somsen Yeah, that wasn't the original paper.
Sjors Provoost Yeah, and then later on there were proposals where you could have different values and then you have to go through a whole bunch of magic to make… so that you can make notes of different values, but it just boils down to having a very large bag of coins.
Ruben Somsen Yeah, well, so it's not necessarily a lot of magic because you can just have the bank have a different key and that would be sufficient so then every key is representative of a value?
Sjors Provoost Yes,
Ruben Somsen so that would be basically enough, but there are lots of different schemes. that's true.
Aaron van Wirdum Anyways, getting back to the main point, the trick is essentially, Sjors, you create a piece of currency, essentially, but then you scramble it with a secret number. You send it to me. I sign it. And the magic is that you can unscramble it while the signature is still valid I can still see that, yep, I signed this, I signed the scrambled version of this, and therefore this is a valid note and when Ruben hands it to me I'll give him a real Bitcoin in exchange in this case or in Chaum's case a dollar.
Sjors Provoost And that means in the case of the bank, they know that, how many, how many notes they issued, so somebody cannot forge more notes, right?
Aaron van Wirdum Yeah, okay, so that's uh, e-cash now we're getting into sort of this new Bitcoin version of it. I think in general it works the same, right? The blinding mechanism and the way transactions work and that's essentially the same. But there's a new aspect to it, which is the federated part of it, right?
Ruben Somsen Yeah.
Sjors Provoost And the fact that you're pegging in and out of Bitcoin.
Aaron van Wirdum Yeah, we're not using dollars or euros anymore. We're using Bitcoin. But I think the interesting part to zoom into now is the federated part. So what does that mean?
Ruben Somsen Yeah. So the federated part is basically that because you can hold Bitcoins with more than one person, you can have the mint or the bank, as we called it, we can have that entity also be some kind of multisig thing, multisig entity of multiple people. So maybe the three of us, we can come together and we can say like, hey, we are going to start Mint. And then anyone who trusts us can deposit some coins and we give them tokens that we sign.
Aaron van Wirdum Yeah, so to take one… To get back to our previous example, I was the bank in the previous example. In this example, I am basically replaced by a multisig of multiple people.
Ruben Somsen Yeah. So let's say it's the three of us, for instance.
Aaron van Wirdum Sure.
Ruben Somsen And then we have a two of three multisig. So for someone to enter into the system, they might send us 100 satoshis. And then we might give them a hundred signatures where each signature is worth one Satoshi, for example.
Sjors Provoost Or we give them one signature that's worth a hundred Satoshis. So there's some flexibility in the denomination. Yeah.
Ruben Somsen Or 10 worth 10, et cetera. Yeah. So when we do that, that signature is also a multisignature of the three of us. So the signature itself…
Aaron van Wirdum So the blind signature.
Ruben Somsen The blind signature, yeah, is also a multisignature. So that means that the three of us, we at least know like, okay, well, it's not just Aaron who suddenly is creating a bunch of these tokens. The three of us will have to have signed off on it or two or three if you want some kind of threshold signature scheme. And… that basically adds to the security where now instead of having to trust one entity, both the currency that's backing it, which is the Bitcoin, is held in a multisig and the issuance of new tokens on this blind e-cash system, they are also a multisig.
Aaron van Wirdum Right. So it's safer. That's essentially the point of doing it federated as opposed to just a single signature solution.
Sjors Provoost And then the question is, what can you actually do with these tokens? Well, if, you know, within the community, people can use it if they trust the, basically the quorum or the, yeah. If they trust a group that signed it, they can use it because they know what these tokens are worth and they can just move them around as it would be with other eCash systems.
Aaron van Wirdum Well, so let's make this very concrete. The three of us have this... We have a multisig mint, the three of us. We're issuing tokens and we're getting Bitcoin in exchange for that.
Sjors Provoost Yep.
Aaron van Wirdum Right? And so the recipient of that token can pay… whoever will accept the token, essentially, right?
Sjors Provoost Yeah, anybody who trusts the token. And that really relies on them trusting the Federation.
Aaron van Wirdum And what's the benefit exactly? Why use this instead of Lightning, for example?
Ruben Somsen Yeah, so I think... One of the things you should compare it to is, I think, a centralized thing, where if you just have full custody, you give full custody over to Coinbase, for instance, or we've seen a bunch of these Lightning wallets that are actually custodial. So you basically have a custodial relationship with one entity. And if that's the case, you can do better with this federation, and you can do better with these anonymous tokens, where if you have an e-cash system, you can at least within all the people, all the people that are trusting the same entity or the same federation, between those people, you can now send anonymously without the federation itself actually knowing where those tokens are going. So that's a big improvement, I would say.
Sjors Provoost Yeah, so it's a more private custodial wallet, but with the same drawbacks as a custodial wallet in that as a user, you don't know if it has full reserve or not. The same goes with a normal custodial wallet. But the advantage is because you're not doing anything on-chain other than making your first deposit and eventually taking coins out, you have very low transaction fees.
Ruben Somsen Yeah.
Sjors Provoost But there's also an obvious downside, which is that, well, you can't use those tokens outside people who trust that particular federation. And the way this is designed, there could be hundreds of different federations. Every family could have their own federation or every country could.
Aaron van Wirdum Well, I think in that sense, the Bitcoin Beach wallet is a pretty good example. I was at Bitcoin Beach for a while in El Salvador and I did use the Bitcoin Beach wallet because it was just a bit easier. And I'm not putting my life savings there because I know it's custodial, but it was definitely the fastest and easiest way to just pay for your breakfast and walk around town and get a Coke.
Sjors Provoost And so in this new situation, you would again deposit your regular Bitcoin from another wallet into the Bitcoin Beach system. But they would then give you these Chaumian eCash tokens so that you can now pay people without them knowing who you're paying.
Aaron van Wirdum Exactly, yeah. The biggest drawback was privacy. Whoever was controlling the funds or managing the wallets could see exactly where I was having breakfast or where I was buying my drinks.
Sjors Provoost And then the next step that is, I think, a pretty smart leap, although it has actually not been completely developed. So we'll have to see if that leap is vaporware or whether it's going to be real. That is that this federation will now make arbitrary Lightning payments for you.
Ruben Somsen Yeah.
Sjors Provoost And that means that if you want to pay somebody on the Lightning network, yeah, you can do that. You'll send your tokens. You'll essentially burn your tokens back into the federation and they'll do the Lightning payment for you. And this is nice because running a Lightning node by yourself is not super easy. So the idea of having a joint lightning node with even just with 10 people makes sense. And you can use these tokens as a way to sort of bridge between that.
Aaron van Wirdum Yeah, the point here is being, so if you own these tokens, so you could withdraw, you could redeem the tokens for Bitcoin and then send the Bitcoin or Lightning. But what you're saying is the trick here would be if you want to pay someone else, you only have these tokens, then there's some sort of trick where the redemption is sort of the same thing as a payment to the other person.
Sjors Provoost Yeah, I mean, redeeming money, like in the original eCash example, you would redeem coins and then your bank would give you a balance. In this case, you would redeem coins and then they make a Lightning payment to the destination. And you can probably use various different schemes depending on how much you want to trust the federation to actually do your Lightning, but you're already trusting them to redeem your coins. So you might as well trust them to make a Lightning payment for you. But this really scales well, right? Because you could also have the federation do a [unclear] payment, but then you might as well just have your own non-custodial wallet because for every payment to the outside world, there is exactly one Bitcoin transaction. So there's no advantage to putting your coins into a federation and then paying from the federation other than maybe some privacy.
Ruben Somsen Yeah.
Sjors Provoost So Lightning does give that advantage of scale where one Lightning node, which might have 10 channels or whatever it is, can handle the outgoing payments for a large group of people. Sharing the overhead.
Ruben Somsen So there is one thing that I should clarify here because there is the theoretical model where the federation controls a Lightning channel and then you can peg out through that channel. But that actually in practice is very complicated because now you would have to have a, you know, Lightning is a two of two multisig, but now one of those two is federation. And that is actually very complicated in practice for the federation to run the channel itself. So actually the solution that Eric is working on now is one where instead of having the federation itself control the channel, he actually suggests we can basically create a token that has some smart contracting features that are very similar to the Lightning Network, how the Lightning Network works itself. So basically, you can open a channel that only has the amount of value that you want to send out on top of the federated eCash system. And then you make that one payment and then your channel is basically empty, so your channel closes again. And that allows basically anyone to sort of play the intermediary, where you could think of it similar to maybe the Liquid network that we've discussed on, I don't know which episode, maybe.
Aaron van Wirdum [unclear]
Sjors Provoost I don't think we have discussed the Liquid network. We've mentioned it, but we haven't done a full episode.
Ruben Somsen Yeah, okay, maybe not a full episode. The Liquid Network is, just for people that know it then, at least, is kind of a separate blockchain that is also controlled by a federation. But anyone could open a Lightning channel on top of Liquid and then connect that to the Bitcoin Lightning Network and then send out of Liquid into the Bitcoin Lightning Network. And this is kind of a similar thing where you do the same thing, except the channel now kind of exists inside of the eCash system.
Sjors Provoost Yeah, but I guess the big picture way of putting this would be, say, well, you could have some sort of hypothetical Lightning node that is run by the whole federation. This is difficult. It'd be nice if one person can run a Lightning node and then interact with the federation by accepting these tokens. And then, you know, there's different ways to do that. And one would be to use smart contracts.
Ruben Somsen Yeah.
Sjors Provoost But in the end, yeah, that's nice.
Ruben Somsen Yeah.
Sjors Provoost And maybe, you know, as a big picture— big picture thing, one of the things with the original DigiCash-like systems or eCash systems is that every merchant has to accept the specific little cash token in order for it to be useful. So you need to win the network effect game. They almost won that because they almost had a deal with Microsoft that would have put DigiCash on every PC out there. So that's pretty cool. But in this case, with this new proposal, you don't need that big of a network effect because as long as you can exit the federation through Lightning, you can interact with anybody who takes Lightning, which already has a much larger network effect than your little federation has. So that means you can do these federations at a much smaller scale.
Ruben Somsen Yeah, and maybe to bring back kind of a practical example, you could have a group of people that, you know, trust each other pretty well and they create a federation and, you know, maybe it's a local community and maybe like Bitcoin Beach. And then really, as long as there's one person within that community who's willing to run a Lightning node and trust both the federation and trust the eCash, as well as that person being connected to the rest of the Lightning network, now you can send out payments out of the Lightning network. And then you can imagine that on the other side, maybe there's yet another eCash system with another person who is doing the same thing. And now you can send from one federation to another federation. And you sort of have this entirely… the system that is sort of connected.
Aaron van Wirdum Ecosystem of federations, yeah. Do we call this layer three? Is this layer three, Ruben?
Ruben Somsen It's a 2.7, I think.
Sjors Provoost And that would be nice if all custodial wallets actually just switched to a system like this so that there'd be good privacy in custodial wallets. Unfortunately, if they're based in most countries, being too private would actually cause some regulatory issues for them, to put it mildly. So they may not be incentivized to offer this type of privacy to their customers. But that's sort of a regulatory battle that is difficult to win. I mean, there's a reason banks don't use Chaumian eCash. I think it's because, not because they can't, it's not because they don't have the developers, it's just because they can't offer privacy to their customers. They don't want to, for regulatory reasons, they don't want to for selling user data reasons. So it really probably boils down to people deciding themselves to set up these federations between friends or other trusted groups.
Ruben Somsen Yeah. And that is, I think, sort of the limitation of the technology, right? Where you are custodial if you run a federation like that. So you can't really grow too big. If you grow too big, it's very likely that you're going to get regulatory scrutiny on you. So that makes it difficult unless you maybe have a multi-country federation, but those are kind of difficult too.
Sjors Provoost And you also run into trust issues, right? Because with three friends, you might still trust that the system is solvent, but with some random people running, you might not trust that.
Aaron van Wirdum Let's zoom into that for a second, because this is one of the main drawbacks, is that you do have to trust the Mint, the Federation in this case. But is that something that could potentially be solved with provable reserves or that kind of…
Ruben Somsen So, yeah, so unfortunately, it's sort of the trade-off of getting this really good privacy. The privacy is basically so good that you can't audit the supply. That's basically how you can see it.
Sjors Provoost Right, because in theory, you could audit the supply by, for example, when you mint a coin, you would mention which UTXO it's minted from. And then if enough people say, hey, I got a coin from this UTXO, and I got a coin from this UTXO, and if that adds up to more than the value of the UTXO, well, then you kind of caught them lying. Essentially. But you don't want to actually be able to link the UTXO to these specific tokens because that defeats the whole privacy point of it.
Ruben Somsen Yeah, it would sort of become basically like a blockchain, right? Where you have a history of transactions and sure, there are pseudonyms and maybe you can have something like confidential transactions to make it more anonymous, but there are still links. So you don't get that perfect privacy. But that also, so that means also that you, because you can't audit the supply, you know, one of the things that I think is also a limitation here is that if the federation wants, they can actually secretly cheat where instead of them just taking all the on-chain Bitcoin and just moving it elsewhere, which would be visible, right? Because you would see the Bitcoins move on the Bitcoin blockchain and you'd be like, hey, that's a lot of Bitcoins are moving. They could also inflate the number of tokens and that would also be invisible. And then basically the e-cash federation would be running a fractional reserve.
Sjors Provoost Yeah.
Aaron van Wirdum One of the things you mentioned, or no, one of the things in the show notes, which I'm not sure if we still need to get into our specific use cases or any specific use cases of this that we need to mention or have we gone over that?
Ruben Somsen I feel like we've covered it, right? And so maybe one thing I can add here is that, you know, I'm seeing like when it clicked for me, why I think eCash is useful and necessary in Bitcoin today is actually because we've been seeing a lot of these custodial Lightning wallets pop up. And they're pretty popular. They're pretty easy to use. So I think it's the perfect niche to like just, as of what is happening right now, I think, yeah, that's just straight up an improvement. Even if the federation is a federation of one, I still think it's an improvement because at least everybody who trusts that single custodian can now anonymously transfer between them.
Sjors Provoost Exactly. Because we've been talking about federations all the time, but it could be an N-of-one federation.
Ruben Somsen Yeah.
Aaron van Wirdum Right. Okay. Sjors, you wanted to break this down now. You wanted to attack the privacy of this system. Is that right?
Sjors Provoost Well, yeah, a little bit. So we mentioned something about denominations, so that we think that the original system had only one denomination. Maybe, was it cent? Maybe, I don't know. But there have been proposals, and the new system that's proposed by Blockstream and folks and Eric has multiple notes, just like your normal cash. You might have a $1 note and a $10 note. And so you can provide change by issuing multiple notes. But in Bitcoin, we have precise amounts. So you might wonder why you wouldn't just issue notes with precise amounts. And there's, I would say, two objections to that. One is that it becomes cryptographically more difficult to do because the bank is blindly signing. And if you want the bank to also not know the amount, well, that's a bit of a problem. So, yeah, and there are some solutions to that, right?
Ruben Somsen Yeah, Brands credentials is one of them.
Aaron van Wirdum The problem with the amounts, and I think it was mentioned, but just to spell this out, is if I issue exactly 1.023, blah, blah, blah. And that exact amount is withdrawn by Ruben later, then I can obviously link the amounts and I still know who paid who, even though the blind signature was used.
Sjors Provoost Yeah, well, that was the second objection I wanted to bring up. So it's more difficult to issue these things in the first place because of the blind signature ritual. And it completely destroys privacy because, yeah, you're sending very recognizable amounts across the wire. So one of the, in general with these eCash systems, you can start, a bank could try to correlate uh, payments to people by carefully watching how much, you know, because every time you receive it, you immediately have to verify it So if money changes hands very frequently, you might be able to infer where the money is moving, though in principle, you don't, you know, in, at first glance, you know nothing because a bank issues coins, so they know how much coins they issued. So Aaron, in this case the bank, you know how much you issued to me, say $20, but you have no idea if and when I spent them because I don't tell you. And Ruben receives these notes. Now, if the three of us are the only ones on the network, then the very fact that Ruben redeems a note, even though he never bought one, means that he got it from me. And then you know that I have 19 and he has one. And this type of thing, if the network is bigger and there's multiple customers, it's… It's much more difficult. It becomes noise. But if you're smart enough with data, you might still be able to infer things. Just like the chain analysis companies out there can monitor the blockchain. And even though they don't know for sure which part is changing, which part is not, they can somewhat infer things.
Ruben Somsen Yeah, just to add a small wrinkle, like even in this case where there's three people and you have 20 for tokens, you sent one to me, even then you could like try to increase the anonymity by just redeeming your own token. Right. So even in that scenario, maybe there could still be some privacy. But yeah.
Sjors Provoost But I mean, the thing is, you know, when Chaum originally proposed these things.
Ruben Somsen Yeah.
Sjors Provoost Things were quite different. This rat race or rat race, arms race, I guess, between the privacy people and the surveillance people was not started yet. So they may have been slightly naive about it. Just like people thought Bitcoin was going to be much more private than it turned out to be.
Ruben Somsen Yeah.
Sjors Provoost And what that rat race would look like, yeah, I could imagine that wallets would start deliberately redeeming their own coins or waiting a little bit longer. And you'd have to think about what the denomination should be if you have multiple denominations. Should you have too many $100 coins? Maybe not, maybe yes. Yeah, um, that sort of stuff.
Ruben Somsen So I think the, uh, the strongest point that, that you made, Sjors, earlier when we were discussing this before the show, uh was that you need a Tor connection, basically, where you, every time you redeem a token, you don't want to connect with the same IP to the, to the mints, because if you do, then you are leaking a lot of information. Whereas if every time you connect to the mints, you, you present yourself as a different IP, as a different person. That really helps a lot, I think.
Sjors Provoost Though I would also say that if custodial wallets allowed you to have totally anonymous accounts with Tor, then we might not even need these type of solutions. So there is the problem of most custodial services not liking the idea of having totally anonymous customers depositing random money. But this, again, might be a matter of scale. If it's just you and your family, it might be okay if you're using Tor. Though then again, maybe you don't need to use Tor if it's just you and your family.
Ruben Somsen That's funny too.
Sjors Provoost So these are interesting questions. And of course, HTTPS was invented since the Chaum thing too. At the time, I think if you were doing this with your bank, your bank could literally see what you were, or your internet provider could literally see what you were doing because internet traffic was not encrypted, whereas now it is. So this already helps a little bit.
Ruben Somsen Yeah.
Sjors Provoost Okay. I think we've covered pretty much what we wanted to do, right?
Ruben Somsen Yeah.
Aaron van Wirdum I mean, the other thing that's in my notes here is ideal properties of digital cash. Did you want to get into that or not?
Sjors Provoost I could very briefly mention this. So there's a paper that Ruben found. I think it's a lecture note from probably 2009 from a UK university somewhere where somebody listed... sort of the top seven wishlist things in a digital cash system. So although this is probably from 2009, it looks like they were not aware of Bitcoin yet.
Aaron van Wirdum And is this all focused on privacy or not necessarily?
Sjors Provoost Well, so digital cash, he would define it as being able to make payments that are indeed private.
Aaron van Wirdum Right.
Sjors Provoost So whereas PayPal is not digital cash, it's more like a digital bank account because the payments are completely traceable.
Ruben Somsen Yeah.
Sjors Provoost And so... He would say, ideally such a system, and he says there is no such system.
Aaron van Wirdum I think that's worth pointing out because throughout the whole block size debate and all of that, still people think that digital cash refers to the cost of transaction, but that was never, at least not the original meaning of digital cash. It was always about privacy.
Sjors Provoost It depends on what you want, right? I mean, In this case, digital cash has the emphasis on cash. And not cash...
Aaron van Wirdum Cash and privacy, Sjors, right?
Sjors Provoost Yeah. Well, yeah, because the cash is private. Others would say...
Aaron van Wirdum That's my point, we agree.
Sjors Provoost Exactly, and others would say, no, it's about being cheap, but being cheap can be done with the database.
Aaron van Wirdum Anyways, what we're referring to, what we're talking about here is privacy.
Sjors Provoost Yeah, so privacy would be secure, so you're able to not fake these coins. Well, that's pretty much solved with digital signatures in general. It should be anonymous, which kind of means like two people, like I pay Ruben and Ruben doesn't know who I am. I don't know who Ruben is. And the bank doesn't know who either of us is. That would be ideal. Now, the system we discussed with the original cash system, that would not be the case. The bank would know who I am. The bank would know who Ruben is, but they kind of don't know that we transferred, that we transacted.
Ruben Somsen Yeah, I mean, not necessarily, right? Like it depends, I guess, on that. If you assume a trust relationship between the bank and every individual person, then yes. But you could also, maybe if the bank is crazy enough to allow anonymous people, then that wouldn't even be necessary.
Sjors Provoost Something like Tether could exist. It's portable. So that means you can use it everywhere. I guess portable is something we take for granted now that we all have mobile phones. So maybe we don't really have to go into this. It's two-way.
Ruben Somsen Well, I mean, one thing I find interesting about the portable one is, so it's portable in the sense, I think that gold would be very heavy to transfer. It would be very difficult to transfer over long distances. And, you know, having a digital cash is like this thing that you can easily transfer. But, you know, one thing that's sort of funny is that I do think for the Bitcoin blockchain, at least, the block size is sort of constrained on portability, you could say. It's not literally portability, but there's a limit to how many people can transfer, uh, their bitcoins, um, and, and obviously if you pay enough you get to transfer your bitcoins, but that is, I think, sort of a portability limit.
Sjors Provoost Well, no system is perfect. So the other thing is they call it two-way, and then the second term is peer-to-peer, and that makes it more clear. So with Visa and MasterCard, you have merchants who sign up with the network as a merchant, and you have customers who sign up as a customer. So two customers cannot send money between their two credit cards. Now, there have actually been some prepaid debit card systems out there in more recent years with fintech companies that do let you essentially use credit cards as a peer-to-peer system. So there were workarounds for it. But this was considered a big problem. And, you know, the eCash system solved that because anybody can send the cash to anybody else. And so does Bitcoin, of course. Everybody can send Bitcoin to everybody else. It should be offline capable. Now, offline… There seems to be more defined, not so much as not having internet, but more about not having access to the third party. So they say, well, you can't make a payment if your e-cash bank is offline even for a second because you can't verify it. And I guess what at least Bitcoin lets you do is… not— both parties don't have to be online at exactly the same time. And the blockchain has to respond at some time, at some point, but it doesn't immediately have to respond because you have to wait six confirmations anyway. So it at least removes the requirement of being online simultaneously. Whereas, you know, a lot of pin systems or terminals in shops, they really have to be on and connected to the bank and every other central party involved in that whole process or the whole thing doesn't work.
Ruben Somsen Yeah, I will add to that, that at least for eCash, there have been attempts to make it literally offline where you can just go to a merchant. And I think it also makes sense at the time because, you know, internet was less widespread than it is today, to literally…
Aaron van Wirdum Your computer was still making a lot of noise in order to get online.
Ruben Somsen Exactly, yeah.
Aaron van Wirdum We're talking about early 80s.
Ruben Somsen Yeah, terrible 28k modem that I had was lovely to listen to.
Sjors Provoost I think we're talking mid-90s.
Ruben Somsen Yes. But yeah, so the idea of actually being able to go to a shop, not be online, give a token and then not have to worry about, like, basically double spending. I think this was solved in kind of a weird way where the token sort of had my name on it, but only if I double spent the token would, would the, my name be revealed, therefore I'm disincentivized from…
Sjors Provoost Yeah, There's two ways to deal with it, right? One is to punish afterwards, yeah, and the other is to make it physically sort of impossible with trusted hardware, except, you know, trusted hardware is a can of worms. Yeah, but that's of course in, in Bitcoin we have the OPENDIME yeah, where you can technically give some money offline, although not really because you can't check the balance without being online anyway. So it's not perfect. And then number six was wide acceptability. So we kind of talked about that. Bitcoin still has, of course, this problem of getting a bigger network, but it is at least globally known and recognized.
Ruben Somsen I think it kind of goes hand to hand with number seven, which I think was user friendliness. And I think the more user-friendly it is, the more widespread it will be. I think those two kind of go together.
Sjors Provoost But I think an interesting aspect of those two, I guess, is that you can have small federations, as we discussed. You can have small federations and still interact with everybody in the world that speaks Lightning. So that's a nice way to avoid this phenomena where Visa and MasterCard, if you want to compete with them, you have a big problem because you need to convince every merchant out there.
Ruben Somsen Yeah.
Sjors Provoost And if you want to compete with Bitcoin, you also have a big problem because now you have to compete with this coin.
Ruben Somsen Yeah. And that's really thanks to the Lightning network. And it's really nice, like a nice feature of the Lightning network that it's, it's sort of, it works on every chain or in this case, not really a chain, but any, any kind of system that has some smart contracting, you can kind of put the Lightning network on there and you can have it be a glue that connects everything, everything together. That's really nice.
Sjors Provoost Yeah, maybe, I mean, I think the Bitcoin itself already gets you there. Yeah, but, and I guess this federation system, you could do it without cryptocurrency too, right? So yeah.
Ruben Somsen It's all, it's all possible.
Aaron van Wirdum Yeah, so that was a pretty long digression, I think.
Ruben Somsen Yeah.
Aaron van Wirdum But interesting, and I was wrong about the privacy part. So this was about all of the privacy, uh, digital cash features, like you said, apparently just how to reproduce cash digitally.
Sjors Provoost Yeah, that's a challenge. And I would say Bitcoin solves it better than eCash did, but the combination might be even better. So that's pretty cool.
Ruben Somsen Yeah, I still think there are nice use cases for all different types of layer two solutions, basically.
Sjors Provoost All right.
Aaron van Wirdum All right. I think that's it. That ties it together for me.
Ruben Somsen Yeah, so I think just one thing to clarify, at least for those who are super interested in the tiny little details, I think. So there is a version of eCash that's called Brands credentials and other variants as well, I think. So those do allow variable amounts without this problem of every amount being its own anonymity set. But that, unfortunately, doesn't really work, or at least not easily. It's not easy to make it work in a federation setting. So that's why it's not being used. Maybe in the future, we'll think of the proper cryptography to allow that to be used. But as of right now, that seems a little complicated.
Sjors Provoost All right. Well, then, in that case, thank you for listening to Bitcoin Explained.
Aaron van Wirdum There you go.