Increased Wallet Support with EIP-7951 in Fusaka

Transcript

This new one gives us some superpowers which are pretty cool. It's as simple as I could have a wallet on my phone where I just use Face ID or whatever. I want to make a transaction, some trade on Uniswap, I just look at my phone, it all verifies, it's all secure and we're off to the races. The reason that this R flavor is so cool is because there's like wide support for it existing hardware environments. I think we'll see good adoption and that's definitely something to keep track of. Welcome to the Fusaka file, a limited episode podcast on Ethereum upgrade.

This is a podcast that focus on enterprise use cases. I am Puja Randan from ECH >> and I'm Paul Brody, the chairman of the Enterprise Ethereum Alliance. In this series, we are trying to unpack the enterprise impact of Ethereum upcoming Fusaka upgrade. So, we are hoping to get more information in plain and simple English. Today, we are diving into a proposal that quietly bridges the world of hardware security [music] and Ethereum smart contract. EIP7951 pre-ompiled for SECP 256 R1 curve support. This proposal also addresses critical security issue discovered in [music] RIP 7212 while maintaining full interface compatibility with existing layer 2 implementation. It may sound

technical but this EIP could open doors to simpler, safer authentication for millions of users and a big step forward for enterprise adoption. Joining us today is ACDC chair Alex Stokes from Ethereum Foundation. Welcome, Alex. >> Thanks for having me. Yeah, I'm excited to be here. Alex, thank you for joining and uh could you please start by introducing yourself? Tell us a bit about your background and how you became involved in Ethereum protocol development and cryptographic research. >> Sure. Yeah, there's a lot to say. I've been in Ethereum for a while now um at the EF for many years and in Ethereum before that. My journey into Ethereum is probably like many of the listeners or at least you know many of the other different uh people I meet in the

community thought it was really interesting. You know, I remember reading the yellow paper a long time ago and it just captivated me in a certain way. And from there I was like, okay, like this Ethereum stuff sounds pretty cool. Like how can I get more involved? And it turns out, you know, something we still have to this day. Uh I think we have great strength of Ethereum as an ecosystem is this emphasis on open source. So uh my background is yeah engineering research some combination of these two. So many years ago now, yeah, I started working I just found some open source projects at that time around working on ETH2 that would ultimately become the beacon chain that we have today. And yeah, I just started contributing where I could and just got in deeper and deeper. One thing led to another and yeah, I don't know how much

we want to go into the history, but I'm here. >> And now it's too late for you to get out, isn't it? >> Yeah, it's all I think about is Ethereum for better or worse, but I'm okay with that. >> Well, thank you for joining us. I want to start by just asking you to talk about this EIP and you know to the maximum extent possible just give the basic thumbnail sketch of what this is and what it's supposed to do and then we'll kind of iterate from there with a few questions about the likely business impact and user impact. >> Yeah, sounds good. So, right, the ZIP the title might look kind of imposing. There's this like very long random string SEC P256R1 and yeah, so how to think about this? So it's referring to a cryptographic construction. Yeah, maybe just to keep

it sort of high level for now, if you go to make a transaction in your wallet on Ethereum today, you'll at some point realize you need to like make a signature. Like this is how we keep the chain secure, you know, if there's like some coins in my wallet versus in your wallet. Ultimately, there's this thing called a private key or a secret key that only I know. And the way the cryptography works is that with this key, I can produce a unique signature. And again from there, the way the chain works is it only takes transactions that have a valid signature. And so what this means then is that you know you let's say I have some USDC in my wallet like you can't spin it on my behalf unless I give you some sort of authority to do so. But today on Ethereum when you go to

do this essentially everything starts from this curve or well maybe I'll take a step back but it starts with a particular type of cryptography we have. It's very analogous to the CIP. it would be SECP 256 K1. And again, tell me if you want to get into the details here. There's a lot of history around K the K flavor versus the R flavor. Why Ethereum uses the K flavor and why we're adding the R flavor. But long story short, this EIP is adding what's called a pre-ompile to use this new type of elliptic curve for this new, you know, it's a very similar similar signature scheme, but it's a different flavor of how the cryptography works. So tell me like why are pre-ompiles useful? Like what is it about a pre-ompile that's worthwhile?

Does this mean the work gets done in advance or like what's actually going on? >> Right. So a cool thing about Ethereum, one of the reasons we're all still here [laughter] is because it is a blockchain that is programmable. Meaning, you know, when you go to think about writing software, you know, in maybe a non-blockchain context, you can write these, you know, you write programs and there's sort of a set of capabilities they can do. And without getting too deep into the theory, basically there's there's a lot of stuff you can do. You can write these like general logic statements so that if you want to like add two numbers or like find a word in a sentence or something like this, there's all this stuff you can do. Once you have data, you can compute on it. And so the EVM in Ethereum gives us this

computational environment and okay so why pre-ompiles? So the reason why is because we have this computational environment and that means you can write programs to anything and so like a signature verification like I was just saying like when I go to make my transaction when we verify the signature this is a computation and today when we do this there's basically native support for this K flavor of the curve we have today not quite yet the R1 this R flavor this is on going into Paka and okay so what does this mean is that you could write a smart contract that verifies this R flavor cryptography today. People have done this. You go to do it and it turns out it just costs way too much gas. You know, I'm sure people familiar

with using Ethereum have run into this concept of gas. We need to meter the operations you do on the blockchain. Otherwise, someone could come in and basically destroy the chain because they make everyone compute some like, you know, crazy thing that takes 10 hours and then the chain halts. So, to get around this, we have a concept of pre-ompiles. And a pre-ompile is a way to kind of do what we do today with these, you know, the signatures on Ethereum today. It becomes native in some sense. So like rather than have to write a smart contract that does this, you delegate to the actual client implementation, the actual like software that maintains or hosts the EVM and yeah, TLDDR you get cost savings because rather than have to like pay gas for every step of the computation, which again is you know like millions of gas

and that becomes too expensive. This now prices it in a way that is much cheaper. So, but it's not just a change in pricing, right? And I try to do a little homework and you have to tell me now if I've understood this correctly. What we're not changing the work that has to be done, but we are changing where and how the work is done. So, if there's no pre-ompiled, the work has to be done on chain at a very high level in in kind of the most expensive way. If it's a pre-ompile though, if we're saying this is now a standard function, then it can get done on the device in the cheapest and fastest way and it gets done in a way that doesn't cost everybody money all the time. And so we're not getting rid of the work. It's not like it's

pre-done. It's just we're changing from expensive location that's slow to cheap and fast location. Is that is that a good am I understood correctly? >> Okay. Now, why do we need this new flavor of ice cream or pre-ompile? Like, what's wrong with the old one or why do we want the new one? >> So, there's nothing wrong with the current signature scheme that we have today. It's secure. It works. It's what's maintaining Ethereum today. This new one gives us some superpowers which are pretty cool. And the reason why it gets into like the history of why we have these different variants of the curve and you know, again, long story short, there are just two different cryptographic schemes. The reason that this R flavor is so cool is because there's like wide support for it in

existing hardware environments. Yeah, maybe to dig into the history a bit. So this R curve is essentially derived from a set of parameters maintained by NIST which is like a standards body in the US. And that's cool. It kind of gives people this like sense of like government approval that it's an official curve in some way. And what that meant is there's kind of a green light to go and put these into all sorts of environments. So even outside the blockchain context, you think about your banks and like different financial institutions, they also have similar types of cryptography, you know, these keys they need to maintain and secure. And so like a common thing you'll see there is like a hardware security module, some type of enclave. And these all have support for this R flavor, but not the K flavor.

>> So that means like your iPhone, your desktop PC, the banks, computers, these all have native support for this kind of signature. And this basically automates and onboards all these potentially billions of devices, >> right? Yeah. And so, yeah, jumping forward a bit from maybe even just a few years ago in history, we now have Yeah, like you're saying, we have, you know, iPhones, all these different mobile devices. There's, you know, fingerprint verification and laptops now. There's all these different types of like biometric authentication that ultimately go back to some similar sort of hardware enclave on the device and these all support the R curve. And so yeah, like you're saying, you know, again, what's the high level impact of the CIP or this feature? Well, it gives you much better

UX because, you know, rather than me need to think about like seed phrases and my wallet and like trying to sign things that way. You know, it's as simple as I could have a wallet on my phone where I just use Face ID or whatever, I want to make a transaction, some trade on Uniswap, I just look at my phone, it all verifies, it's all secure, and we're off to the races. So wider access, simpler access, better usability, but also potentially computationally much much cheaper than the prior alternatives. Who did we have on in the last episode? Pri was talking about the goal for Ethereum is to be able to do 100 transactions per person on Earth every single day.

>> He was Justin Drake. >> Justin Drake. Okay. So we were talking with Justin Drake and he he laid out this kind of megaized vision of like hundreds of billions of transactions. And uh one of the things that that I was thinking about when you were talking about like better faster, how this simplifies things is this old like Steve Jobs story about how you know he asked the developers like hey you got to reduce the boot up time and they're like why? And he's like well what if we save 1 second for every person on earth what's that worth right? And the answer is a lot. And I sort of think about okay what if we say like if we think about a blockchain future which is literally hundreds of billions of transactions a day that means we like every transaction gets a signature right. So we need to

save hundreds of billions of computational cycles in order for the Ethereum to scale up at that level. >> Definitely >> got it. I think that sort of gives me that perspective. Will this like practically speaking I build an app I have high volume transactions. Is this going to reduce my gas fees like in a meaningful way? >> Yeah, hard to say. So, the pre-ompile should be priced similar to like a signature today with this K flavor of the curve. So, I don't think it'll necessly be cheaper, but you'll have again much better UX for all the reasons you know that we were just talking about.

>> So, if I switch from a non-p pre-ompile to a pre-ompile, it has a big impact. But if I'm just expanding the range of devices, I've widened my market. I haven't necessarily reduced my typical transaction cost but I made it accessible to all kinds of new users in a much simpler way. >> Yeah, for this EIP there's other things we can talk about around batching and you know moving to rollups. There's like many other ways in which costs are being saved but just in isolation this EIP is more about unlocking new UX and usability security things like this not so much directly targeting cheaper transactions by itself. So this will affect like if I'm an enterprise, this does not probably affect my back office operations because I probably already had pretty sophisticated infrastructure

and I have sophisticated developers who don't care that much about UX. But it does impact my ability to easily onboard new users as consumers who can use touch ID, face ID, fingerprint scanners and others and traditional like username, password to authenticate on their personal devices. >> I think so. And I would also call out that you know it's not so yes for existing you know players but also for new players like I think because this again expands the scope to what type of environments are accessible it should lower the barrier to entry for more people to join right there might be some you know bank somewhere that for regulatory reasons they couldn't use this other type of cryptography there's you know very there's a lot of

legislation around what's what you can use here in which context and the cool thing about this EIP is it's using it's adding this new curve again like a cheap accessible way that you know is more compatible with some of these more legacy contexts. So you could imagine, you know, there are banks now, you know, there's been this huge, you know, uptick in interest around stable coins this year. Many people are thinking about this. They want to get into see where they can enter the game and uh yeah, this should basically make it easier for them. >> Were there particular industries or users who are like specifically calling for this like what's the or tell us a little bit about the origin? You know, no EIP sort of emerges from nowhere.

Somebody somewhere thought to themselves, I have a reason why I want this. So maybe is there an origin story for this EIP? Yeah, definitely. I don't know if there's like a concrete, you know, single origin story, but many people have been asking this question for a very long time in Ethereum because again, this is a very sort of standard, well supported curve in many, many contexts outside of what a crypto native might find. And so, there's always been the question like, okay, like, yeah, why do we not have this? And especially again I think over the last couple years with the proliferation of different hardware you know elements on phones and things people moving more to like mobile first environments you get to this place where you just really yeah you can't help ask the question like why why aren't we using this and it's coming in

for Saka so very exciting >> that's fantastic puja do you want to add some questions here >> yes I would be curious to hear thoughts on like what enterprise people have to do there are certain products which are already using this our version of cryptography Raphy, do they need to do anything in specific to leverage what is coming up with Fusaka? >> Yeah, so it kind of depends on what they're doing. Like if you want to make a consumerf facing app that you know takes advantage of this in a wallet, then you'd need to update your wallet software to do so. One thing I guess to call out is that you know this adds the functionality to the protocol, but there's still a few layers on top of that that need to sort of be sorted out.

So you'll need to see wallet adoption, which you know will likely take some time. Then I think that's more on sort of the consumer side. If you want to think about, you know, enterprises by themselves, if you're an institution that's custoing all that crypto, you might want to look into this simply because now I think yeah, like the point here is that you can now use different types of hardware, you know, environments that you probably are used to using in other contexts, but now you can use them here. So you might want to think about yeah, seeing what it enables. >> And we have also heard about account abstraction like being in Ethereum. What do you think this proposal will also enable that in a better way? Yeah. So there's an interesting, you know, sort of synergy here. I'll say because like you're saying, this is a pre-ompile.

It's not like a native curve to the chain. So it's not that you can go and make a transaction say on your, you know, L1 or roll up today that can directly use this thing. You kind of have to go indirect with the signature verification. And what that looks like today is using some type of account abstraction to take advantage of this. So whether that's 437, 7702, you know, different things like this even like, you know, you could have other things like safe as a multisig like there are many different types of patterns here other than just 4327 or 7702. Point being is you need some type of smart wallet to use this because there's no way to directly access it in sort of the native flow. So yeah, you'll upgrade. If anything, you know, to the extent people want to use this, they will end up using

smart wallet some sort of account abstraction even without knowing it. And you know, that's kind of the vision here. Like I've seen simple demos where it's like you can even have an embedded wallet in some DAP like you don't even really realize what's happening. It's more just like hey I'm here on this you know this application this web page this mobile app. I want to make a trade and it basically just works out of the box again with like biometric authentication on your phone or you know your computer or whatever. And yeah it's it's super cool. >> So that's a step forward. Yeah. Based on what you've seen with like past EIPs, how long should we expect it will take before wallet makers have typically implemented and aggregated this? Right.

So, it's going to go live sometime in December and then how long before you'll see people like this will be a standard thing that will just be routine. >> Yeah, I think it will take a few months at least. So, you know, not only does the feature need to be live, but then you know, wallet developers will need to go and like, you know, understand how to integrate the feature into their application into their wallet. Along with that they'll want to do like security and different reviewing processes to make sure that their implementation is secure. So you know it'll take a few months but I expect yeah well in you know at some point next year we'll we'll have this functionality pretty widely distributed. Yeah, I think this is something that I sometimes struggle to convey to people in this space, which is in blockchain when we

add roll out software updates like this, we're not just making like little feature changes like this is the wallet, this controls your money. And so what ends up happening is a relatively slower update cycle because the stakes are quite high with a sort of unreoverable kinds of issues. And and sometimes people like, but there's this new curve, this new algorithm. We encounter this issue all the time. And the answer is we you can't just swap them out and and be arbitrary about this stuff. It it it requires a fairly deliberate amount of of design and testing and it's always going to lag the technical state-of-the-art because it it requires that time and effort.

>> You want to make sure it's secure. That's the first thing. >> Security is definitely important. And we are also talking about interoperability. In the beginning, I mentioned that this proposal also overcomes some of the uh vulnerabilities that were found with RIP 7212. Maybe if you could briefly talk about it because I know many of layer 2 projects out there would want to learn what this will change. >> Sure. Yeah. So there was some adoption I think at least on a few roll-ups of uh essentially yeah this this other variant. The vision here was you know L2s are not bound by the same sort of rigidity I suppose of the layer one meaning that like you know to change layer one we have to go through the same

process Paul was talking about about making sure things are secure and well tested and you know there won't be a negative impact layer 2 maybe have a little more permissive environment here because you can just kind of roll them up and if there is a fault that is bad but it won't impact you know the entire foundation of Ethereum so because of this you know at least for some roll-ups you can imagine they want to be a little more risky a little more aggressive here and try to pull in new features sooner rather than later. So, some of them did this with a you know different variant of this EIP again as we got deeper into this. Originally, we were looking at just taking the the RIP version and just putting that at layer 1. Turns out there was a security issue and I believe it

was something around like um essentially an edge case around having like either a zero public key or secret key. Either way, I forget the details and we don't need to get into like too deep in the technicals, but ultimately there was a security issue um which we needed to patch and so that's why there's a different EIP now. And yeah, that's just how it goes. We don't want to ship the bug at at layer 1. So that's what we have, >> right? Layer 2 is generally just even when they want to be cautious, they have a lot fewer decision makers they need to consult if they're relatively centralized. Whereas Ethereum's got a pretty like is is you know by design decentralized which means that just the decision to do something alone is always going to be more time consuming. Right?

It's not for nothing that Pua's organization is called the cat herders. >> It's a great name. Yeah. This is what I spend a lot of my time these days at the thinking about and I can tell you it's uh yeah it's way harder than it sounds. >> Yeah that's pretty interesting. Of course there are people to support the ecosystem and we get it especially in Ethereum space. So, we are so thankful for this beautiful ecosystem. Well, my final question for Alex here is I'm just trying to zoom out. What do you see as a long-term impact of EIP7951? Do you have any kind of metrics in mind that perhaps team can look into after the upgrade is done, >> right? So, you know, if we just put on our protocol developer hats, then we can think about, you know, how many people

are calling this pre-ompile and what does that look like. I think something that might be a little more easier to digest is rather than just look at the actual raw chain is just to think about, you know, what does wallet support for this look like? So, you know, that's something I'm interested in personally is six months from now, do wallets offer this feature as as an option or not? And yeah, from there then it's becomes also a little tricky to track each wallet and think about, you know, sort of the ecosystem of them and who's where, but we are doing some of that work at the F at least to help like steward that process. So yeah, I think we'll see good adoption and that's definitely something to keep track of is just yeah, which wallets are are enabled.

>> Awesome. Well, I think we're going to wrap it up here for today. Alex, thank you so much for making your time available today and for giving us this really good sort of deep dive into a feature I hope really does help Ethereum scale into the hundreds of billions of transactions. So, thank you. Of >> course. Yeah, thanks for having me. >> Dear listeners, thank you for tuning in. We hope today's episode helped you understand that despite being a small EIP, this proposal adds to a major move towards bridging enterprisegrade security with Ethereum's open permissionless world. Follow EC institute and EA for more on Fusaka file. Until next time, keep building on Ethereum. Cheers everyone.