On this episode, we’re joined by Andre Cronje, Director at Fantom Foundation. Fantom develops high-performance, scalable and secure technology that supports a wide range of decentralised applications (dApps) and services on the Fantom Network.
- Some of the stigmas and challenges surrounding the DeFi community.
- The need for DeFi culture to mature and be taken more seriously.
- Andre’s role at Fantom Foundation and the types of projects he's been working on.
- The difference between trying to regulate cryptocurrencies and regulating crypto service providers.
- Some of the barriers to the adoption of the decentralised web.
- Specific use cases aligned with Fantom's capabilities.
- Details around how Fantom increased its transaction and consensus speeds.
- The blockchain tech improvements that are necessary for the decentralised web to really flourish.
- The importance of gas monetization for dApp developers.
Thanks for listening to Waves in the Finoverse, brought to you by Finoverse. If you enjoyed this episode, please leave a review to help get the word out about the show. And be sure to subscribe so you never miss another insightful conversation.
Walter Jennings: In the world of DeFi, where is the ideal point of balance between regulation and autonomy? As a prolific developer in this space, having founded several high profile DeFi protocols and describing himself to be a lab accident away from becoming a super villain, today's guest can speak with authority to this topic, none other than the Co-Founder and Director of Fantom Foundation, Andre Cronje. In this episode, we unpack what he really means when he suggests a need to regulate DeFi, the challenges therein, what's possible and useful and what isn't. Andre also speaks about the importance of taking ourselves seriously within the DeFi space to shift narratives away from negative ideas towards more positive outcomes. Welcome to the Finoverse Andre Cronje, the godfather of DeFi. I'm pleased to be joined today on Waves in the Finoverse by Andre Cronje. He is the director at Fantom Foundation and is also known as the Godfather of DeFi. The description I like best though Andre, is on your LinkedIn profile where you say you're a lab accident away from becoming a super villain. Tell me about some of the Appalachian ‘Godfather of DeFi.’ How did that come about and tell me about your earlier days in the development of the summer of DeFi.
Andre Cronje: Yeah, look, it's a moniker I'm not super proud to wear, I don't actually consider myself the Godfather of DeFi. I think that moniker belongs to Robert, Robert Leshner from Compound. It belongs to Kane Warwick from Synthetix and it belongs to Hayden Adams from Uniswap. Those were the three core primitives and Uniswap, obviously the first real AMM; Compound, the first real lender, and then Synthetix, the trade engine and exchange and all those things aside, they were the ones that invented yield farming. They were the first ones to incentivise liquidity by providing the SNX token. So I think they, the three of them, deserve it much more than me.
Walter Jennings: Sorry, I was just going to ask you the ability to develop some of the protocols and the ability to make them really sing. It does appear though that from your time at Yearn, I mean it went from eight and a half, 8 million US in July, 2020 up to 7 billion about 18 months later. You know, you've clearly, your test in production approach to coding has worked well. How were those heavy days of the growth of Yearn?
Andre Cronje: Right place, right time. The markets were going crazy already. The COMP token had launched about a month, two months before that and it had really kicked off the craze of money starting to flood in. And you know, the whole world's economy was propped up to the point where there was just in quotes, free money available that was very easy to quickly accumulate. So just one sort of sideline statement there, you know, test and prod is also something I've really come to hate. What it has become and what it's changed into in terms of meaning. The reason I originally put that up there is because I am a firm believer of personal accountability, personal due diligence, personal sovereign rights. And that's a big reason why I was originally attracted to crypto because it's your wallet, it's your custody. If you lose it, you have no one but yourself to blame. But at the same time, you need to do the due diligence. You need to research these protocols, you need to take care. Test in prod. I didn't test in prod. I had local environments that I tested in. We used test nets, we got audits. You know, we went through all of the tick boxes, but I didn't wanna publicise any of that stuff because the goal was to make sure that the people using these protocols don't have a false level of comfort. Because no amount of testing, no amount of auditing can prepare a protocol really. And we've seen that happen a lot of times. We have protocols that have some of the highest security that have been vulnerable to exploits. So the test and prod was supposed to be a warning, you know. A there be dragons here, trade carefully so that people will take the necessary amount of caution before they interact with these systems so that they are then comfortable with any consequences that might occur. And you know, it ended up meaning the exact opposite. People ended up meaning it to be just the gin into these protocols and just throw your money around and you don't have to test and do whatever the hell you want. So it actually ended up becoming the antithesis of the reason why I stated it, which is sad. And that's also why I don't like the moniker, the Godfather of DeFi. I'm very proud of what I built, I'm very proud of Yearn, I'm very proud of Keeper. I'm very proud of Solidly, the new flat curve AMM there. Some of the best code I've written in my life. What I don't like, you know, is that DeFi, decentralised finance as a general rule is seen as a very degenerate culture thing. You know, it's seen as a, it's a place in the corners of the web where the anons and the hackers and the scammers hang out. That's again why I'm not really a fan of that specific title because I'm proud of what we built and I'm very proud of what it was back in 2018, 2019. I'm not proud of what it's become. And you know, that's why later on in my journey I also started calling for regulated DeFi versus crypto regulation, which we can get into a little bit later or we can unpack now. But in its current iteration it is only ever going to be seen as a bad thing. If you look at something like the Tor Network, right, like technologically speaking, it's great. And everyone uses VPNs nowaday, which is essentially privatised Tor networks. But anyone that talks about the Tor network, you talk about that and it always has a little asterisk. That's where the scammers and the hackers and the exploiters and things hang out. If you're a well abiding citizen, you won't use that to serve the internet. And it's the same if you're a well abiding citizen, you won't put your money into decentralised finance because then there's automatically assumption, oh what do you have to hide? Why won't a bank take your money? So it's seen as lesser when it really isn't the case. And a lot of that is sort of self-fulfilling, no, not self-fulfilling, that's not the right word. But crypto communities did it to themselves. Like every new token was made derogatory as some fruit or farm or animal or like it went out of its way not to be taken seriously. And every one of photo was the fun in games at the time damage has been done. Now if any serious entity looks externally, they see a bunch of children playing in the kiddie pool, they don't see a new technological advancement that can reshape our financial infrastructure.
Walter Jennings: Well it is a challenge Andre, because we spend a lot of time at that intersection where DeFi meets centralised finance or CeFi, and there are so many of the reserve banks and settlement banks and others that are looking at ways to put some of these on the chain. And so there is great interest and there is institutional adoption of DeFi. However it appears that there are serious challenges with the culture and the community.
Andre Cronje: Yeah, massively so. And that is keeping us back. Because if you are a insert random big bank and you know you want to deploy a hundred billion, but you are deploying that along with a bunch of anonymous actors and weird fruit protocols and like silly games and stuff, then is that a place you feel is worth the risk or is that a place you feel you can trust? Sure they can trust the code, sure, they can trust the blockchain but they, 9 out of 10 times aren't going to do due diligence on the code. They're going to check the narrative as it is in the news and in the media and all of these things. And the narrative is bad, the narrative is really bad. And the times DeFi is in the news or crypto really is very likely not the good things. It's mostly because a big exploit or some large hacker group used it to move funds or it's being used to avoid extradition treaties or is used to like launder money, like it's the bad, I feel has far outweighed the good for a very long time. And now, can't really blame crypto people and crypto ethos and all of those things. That is how traditional media decides to portray and that's obviously what they want to focus on, but we aren't helping ourselves. And if you look at, even in the crypto communities, crypto Twitter especially, the speed and ferocity at which these communities latch onto false narratives is insane. I speak to so many project founders that say their biggest stressor is needing to constantly be monitoring what's happening on Twitter because you never know when someone's just going to say something, anything. It doesn't even matter. It can have no substantiating proof, but they just need to say something that sounds sort of sensationalist enough so that it gets picked up by one of these big accounts and it's just a retweet. And then if you confront them it's often like, oh but before it was funny. We're just memeing, we're just joking. But meanwhile they're doing real damage to these projects because we don't self-regulate, we don't self-manage like within our own community, we don't take ourselves seriously. So how the hell is that going to ever change if internally it's not addressed versus and we're expecting people to externally take us seriously, it's just not gonna happen.
Walter Jennings: So Andre, you announced a new job position in April of this year, which was a Vice President of Beams at Fantom. Congratulations on that role. That might be a title you made up yourself.
Andre Cronje: Yeah, look that was a transitionary phase. I had always been with Fantom and the team had asked me to take a public step forward again. You're not going to get me away from the tech but you can get me away from the publicity that's not something that you know sets well with me and I don't handle it well plainly speaking. So I decided I needed a sabbatical away from it and I needed a clean cut away. And then Kong actually asked me, Hey do I mind updating my stuff again so that it just shows that I'm still involved with the project and I'm still there because they're getting tired of telling people, hey I am still there privately, I'm just not there publicly. And at that time we were formalising my role because I’d always just like a founder isn't really a real role. Like that's, that's how you started but that doesn't describe what you're doing in a day-to-day. Me and Kong are Co-CEOs but I mean he deserves that title officially because he does a lot more and I'd been doing some of the architect stuff so he thought about that. Eventually we just settled on adding me to the board and the semi managing director but CEO still having the majority of the duties and I needed something to come back with. But I also figured given the way I left it needed to be a little bit lighthearted. So that's kind of what I was trying to do there. That's like directly against what I just said previously where like we don't take ourselves seriously so how are external people going to take us seriously. But you know there's a lot of Fortune 500 CEOs and stuff that also tweet the most random crap on Twitter that’s definitely not serious and that seems to work out, well or bad, I’m not actually sure. It's just it's when you're involved in a public company it's very difficult for people to separate you, the person and your own opinion and you doing your own stuff, versus you the employee that has a specific role and does specific things. And I think that separation is sometimes struggling as well. That title was me trying to have some fun on my personal LinkedIn. But I do feel I did it at the expense of Fantom. That was me personally conflicting with me employee. Because you know like you're, this is the hotel California, you're allowed to come in but you're never allowed to leave, you're not allowed a holiday, you're not allowed a break, you're not allowed a weekend off because of the hyper public and constant nature. You have to be always online, always communicating, always present, always available and it gets exhausting because you can't share your personal life anymore. You can only be wearing this corporate hat. Yeah, look in hindsight I probably wouldn't have done it now if I had to announce, but I'm still messing around sometimes. But 9 out of 10 times it's always ill perceived so I probably should stop. Like I used to be very opinionated on my Twitter for example as well, but now you'll notice there's no opinions anymore because any opinion I have, even if it's in my personal capacity is all of a sudden levied against the company. I'm like that's not fair at all.
Walter Jennings: Right. Well Andre, I'm not gonna open up the Pandora's box of the challenges you've faced in the past but you have left DeFi and come back to it. Tell us about your role at Fantom and what enticed you to be public about your continued participation in the chain.
Andre Cronje: Look, Michael Kong asked me, I agreed, I have a very long history with him. I have immense respect for him and me and him have been through some real wars. We have real battle scars. There's been some nasty trenches we've been in together. So like when he asked, it was a very easy yes I'll do it if he thinks it's in the best interest of the company. And I do think it's worked out. I'm definitely not back in DeFi. I don't think decentralised finance is something that I would like to be associated with currently. Reason being is I don't like the narrative around it. I don't like what people have turned it into. I don't like the carefree nature of it. It was meant to be financial instruments that give control to the users and really it's not anymore. Like so many of these protocols are just multisigs or team controlled or admin access, very few of them are decentralised. They're basically pseudo banks or companies that are trying to pretend like they're not, while they take user deposits and manage these assets and like often lend it out to third parties and things like that. But the users also don't care. Like they're not gonna put their funds in Compound, which is a lot safer. They're going to go throw it into the next highest yield Ponzi because 99% of the people we've attracted lately with anonymous culture, with NFT culture, with the den culture, these aren't people that care about the original ethos on why we built these rails. You know, they only care about, hey how can I get rich quick? And like we saw something similar happen in early traditional finance as well where it was very boring and then it very quickly became just gambling and then it attracted a lot of the wrong people and a lot of scammers and a lot of get rich quick, and we saw all of those things of explode and I'm talking early 1930s level stuff here now. And after that the regulators stepped in, they decided hey we need to clean this up. The average user on the street won't be able to know the difference. They created some good legislation, some bad legislation. I still think fractional reserve banking was a mistake from the start and I think treasury bills were a mistake from the start. That's different discussions. Other than those two key things, I think overall they've done a good job and I'd like to see them bring that into our ecosystem as well because I do think it needs to grow up and it's very immature currently and not in a good way. So anyway, sorry long deviation there.
Walter Jennings: I do want to dive into depth in a moment about Fantom, but while we're here just for the moment, you had been speaking publicly about the need to regulate DeFi. Are there countries that you see that are building the right framework and how do you regulate a global decentralised industry?
Andre Cronje: So therein I always tried to push that there are two different blocks really or verticals. The one is crypto regulation and that is impossible. No entity in the world is going to be able to regulate crypto. You cannot enforce rules on Bitcoin, you cannot enforce rules on Ethereum. It has no jurisdiction, which means there is no executing entity that can execute the thing that the jurisdictional power is trying to force on them. And them trying to do that just leads to friction. The Canadian government can subpoena Vitalik to switch off Ethereum. Nothing that he tries to do other than pleading, will lead to it being switched off because no single person has that control. But if the regulator's unwilling to understand that nuance, that differentiation, they're going to keep making Vitalik’s work life worse and worse and worse and threatening jail time and eventually go chuck 'em in a jail cell. But there's nothing we can do about it. So if we don't educate and if they're not willing to learn and it seems they've taken that we're not willing to learn stance currently, then it's going to cause a lot of friction for the developers. But the protocols aren't going to change. That's crypto regulation and no amount of legislation can do anything there. The other side of it is regulate is what I like to call regulated crypto. Now there's a lot of TradFi, Crypto intersections. It is, hey I wanna set up a ETF fund for Bitcoin. That is something that can very easily be regulated. You can define structures around that and you can start treating Bitcoin like a commodity or whatever the hell you want to. I think it's a new asset class but whatever fits. You can have an exchange. Again, that is something that takes custody in a personal level, that is something that is easy to create a regulated framework for. A VC. Very easy to create a framework for. There are a lot of touchpoints that you can create regulated crypto frameworks and if you are a DeFi in quotes protocol and you don't wanna be anon and you don't wanna do dodgy stuff in the background and you actually want to have a company and your company is a crypto lender and you want to be registered, you want to be licensed, you want to be able to provide money markets, you're doing it on chain, it's all smart contracts, it's not off chain. You want to have it a legitimate company, then you should also be able to fall in this regulated crypto. You should be able to go create a company or LLC or whatever and you should be able to register and get the licence and you should be able to provide the service. So that's the part I'm looking forward to. That's the part I'm excited about. Because right now you're forced to be anonymous now. Now not all anons are bad, but the anonymous culture does allow bad elements to come in a lot easier. So the problem is now we have to all be anonymous, we have to all hide away or otherwise you're going to get a notice from mostly the US because they're the only ones that seem to think they control everything, and you have to hide. If everyone has to be anonymous, it's a lot easier for the scammers and the grifters and thing to come out of that because they're not willing to put their face behind it because it's a lot easier to steal money if people can't find you. You know, that also makes the space worse because now it's the barrier to entry for scammers and bad actors are so much lower that further causes issues. So all of these things they like spiral on top of each other verses like Compound. There's no reason Robert wouldn't want to go get a lending licence. He's a known actor, his company is registered in the US. Uniswap, same thing Uniswap Labs give us some kind of broker licence or something that fits with an automated market maker. Since we don't actually take assets. All of these are normal incorporated companies with normal known employees with money that they raised with operating margins, with revenue. The thing they interact with just happens to be this technology called crypto. Just by virtue of them using a wrapped B2C versus using a futures contract on wheat, you know it doesn't really make any bloody difference. It's the same thing. So the same rules should apply and I think that's the big difference. What a lot of people have historically thought I meant was hey bring in the regulators, come regulate Bitcoin, come regulate Ethereum, all of that stuff. It's impossible. It can't happen and it just creates frictions on the dev sites. But at the same time we can't just be like, hey screw off regulators, we're going to do our own thing. We have to be helping them to understand, hey, this thing you're trying to do, you can't, but here's what you can do and here's why it's useful. So an example there is the new EU bill. It's a great step, it's in the right direction, but there are a few clauses that are contentious. You know like needing to have proxy upgradable patterns on all of your contracts. Needing to have access control on all of your contracts. A lot of these things completely take away from the protections given via smart contracts. You know, the fact that it can't be upgraded means that I know if I put something in there, it's gonna do exactly what I saw it's going to do. But if I know the team can just upgrade it at any time, then that's a bigger risk factor for me and that actually removes one of the benefits I'm receiving. Those are minutiae points in an otherwise intent driven good bull. Swiss DLT Act. Another great one in terms of being able to tokenize real world assets into crypto while holding it in Swiss regulated entities. South Korea has opened up in terms of allowing security tokens and being able to take traditional securities and to digitise them on chain. So Asia's taken a good stance. China has brought friendly but asterisks regulation, it's like, hey yes we like crypto, it's our crypto. It's very specific. It's like their CBDC, they call it a CBDC but it's not on chain, it's not a crypto, it's a hundred percent controlled by their servers and things and you can't do anything external with it.
Walter Jennings: And for guests who don't know that central bank digital currency, CBDC. Here in Hong Kong we have two exchanges that have been licensed for security token offerings and I think that would be RWA tokenization, Real World Asset tokenization. Yet despite having two exchanges with those licences have yet to see an offering made available or put up for public sale in any of these exchanges.
Andre Cronje: Yeah, look. But like a lot of it is, and this is something I have to tell founders quite a lot, is the decentralised web and you know, we're not building blockchains, we're building a decentralised web. It's a decentralised Internet where the business models change and the ownership rights change and that's important. But it is following, let's say 80%, the exact same trajectory as normal Internet did. So it comes with this problem that the tech moves slow, the regulations and the lawmakers move even slower. But so you have the slow moving tech where we can already see what it's supposed to be because of the Internet. So we can already see, hey where's my Twitter? Where's my Instagram, where's my TikTok? Where's all this stuff I'm used to? Where's my instant payments? Where's my Venmo? Whatever the hell it is, why don't I have this yet? And then, okay, but this technology is broken because I don't have this stuff yet. My argument is always take all of these things you just said, you want Twitter, Venmo, whatever. Now go back to the 1980s and build that app. You can't, because the technology didn't allow it, the frameworks didn't allow it, legislation didn't allow it and it just wasn't possible to be built back then. The technology has to get to a certain point until it is possible to build these things. But now the problem is the tech isn't there yet, but the collective consciousness is already there. So it expects us to already be there and trying to front run/hyper speed rush through all of this stuff. And you know, that's why on the STO offering for example, people that have gone through all of the effort of being a security, they've already applied, they've gotten their licence, they are a regulated security, they're on Hong Kong Stock Exchange, New York Stock Exchange, whatever stock exchange it is. So to them, they're like why do we need access to the crypto stuff? We have more liquidity where we are and there's all the access to the people we want and we know it's legit. So why do we want the extra headache of this external system? And that's where we're at now. At some point liquidity is moving more and more away from these sort of centralised pots and more and more into self custody. People also want more access to something they feel they can control, you know, and we always go through these fear cycles and we're in one of those fear cycles right now. Like the banks thing in the US and Switzerland et cetera is the example there. Because you know now everyone's withdrawing their cash because they're realising hey the cash that's in the bank is an IOU, it's not actually my cash. Like I'm actually lending my money to the bank and if they default you're screwed, you're not going to get your money back. So they rather want to have the cash underneath their mattress. That's the safe space to have it every so often. This happens 1930s, 1960s, 1980s, 2000s. Like it always repeats until long enough has passed that people forget and then they go put it back in because they need it to grow and they need access and it's easier and more convenient. But for the first time we have a parallel system, which is the decentralised financial system which now allows them to, you know, have it themselves but also be able to get those interest rates and those same benefits and being able to lend it out themselves in a pool they can monitor. So there will be a slow transition over time and eventually it's gonna be a lot more attractive for the STO guy to launch it on the STO platform so he has it digitised because by then real world asset systems have been implemented that allows easy borrowing against that asset or allows me to go lend it out or allows me to leverage against or whatever the hell it is I do. Which you know right now is difficult. Like I have to go officially request a stock certificate and then go find a custodian that's willing to accept that and then try and do something against it. Or I need to go sign up for their broker account and I can't transfer the share so I have to sell it that side and then buy it that side. So there's a lot of convenience coming and we can see it, but the underlying, the base layer blockchain isn't there yet. And then even if it was the regulatory side isn't there yet and the regulatory side is 30% of the job because even when the framework is there, it needs external parties, it needs auditors, it needs the PwC, whatever, KPMG, all of these people in the world, it needs them to now also understand this law because they now need to be able to go to, let's take the DLT Act, the RWA stuff in Switzerland. Sure you're allowed to now take a equity amended on chain, but you need to apply certain compliance rules and you need to submit reports and things. Now to get those, you're gonna need a third party audit firm. And now even though the rule is there and even though the technology allows it, none of these auditors are capable of doing anything because right now they're not spending time and understanding it and they don't have big enough inbound requests to really care about it. So first there needs to be a lot of demand for them to then go train and then being able to offer it. There's all of these extra hops we have to still go through but we are going through them, it's just a matter of time. The problem is as said at the beginning, because of this collective consciousness, we can already see where it's going to be three years, four years, five years from now. And now it's just really frustrating trying to wait because these things just take time.
Walter Jennings: Now Andre, we've talked a lot about DeFi but we haven't really taken a deep dive into Fantom, a layer one blockchain and you have had a long history with Fantom. Tell me what are some of the use cases that are perfectly aligned for Fantom's capabilities.
Andre Cronje: Like inadvertently this whole talk, I've actually been talking about Fantom the whole time even though I didn't mention it because you see the thing is most other blockchains have hit a technological wall, and that is due to the current constraints of hardware and CPU and memory and bandwidth and all of these things. And then they have decided, okay, we can't continue pushing the speed of our blockchain, so instead we're going to go for these isolated islands of blockchain. It doesn't matter if you call it L2’s, super nets, subnets, whatever your nomenclature is. What you're doing is you're taking Ethereum, you're copy and pasting it into Ethereum 2, and you're now saying okay but we just scaled Ethereum 1, because you can now do twice the amount of throughput because there's Ethereum 1 and Ethereum 2. That's literally what people are doing with layer two’s currently. So that's taking a very easy way out of the scaling problem, but trying to sell it like it is fixing the scaling problem when it doesn't. If you have funds, if you have a 100 USDC on ETH, it doesn't exist on Arbitrum, Optimism or any of these other L2s. You have to bridge it over and then you have to do things there and all of these fraud proofs and things or zero knowledge proofs that you have to sort of contest if something went wrong is invalidated by the fact that there are third party bridges that have integrated to this thing. Because if it's not gonna go through the canonical bridge and I can just withdraw the funds via something else, your proof that I stole from you doesn't matter cause I already have my funds exited, all you're going to do is hurt the bridge. So there's a lot of practical stuff missing there. So at Fantom we've always taken the approach of hey, these limits aren't done yet. Like we can see they're not done yet. But not a lot of people are taking a performance engineering approach to this problem. You know, one of the first things when Fantom launched, there was only a really proof of work and we released a BFT, which is our asynchronous BFT system, which even in today's world where we have proof of stay competitors and other consensus mechanisms are still the fastest. We have the lowest time to finality of any network. Sub 900 milliseconds until you send a transaction, until it is confirmed by the network there because it's asynchronous BFT more specifically because it's BFT, it has true finalities. There's no longest chain rule, there's no hey please wait 10 blocks confirmations before your transaction is finalised. When it's that 900 milliseconds is done, it's finalised. So consensus was the first bottleneck we noticed because if you looked at the traditional proof of work systems back then, you know Ethereum, Bitcoin, proof of work is fixed time throughput. So you know Bitcoin 30 minutes or 10 minutes or Ethereum, what was it? I think it was 14, 15 seconds back then. So it can never go faster than that. Now you can increase the size of the blocks and add more and more transactions and thereby get throughput in quotes. But that then means you need beefier and beefier hardware and that just becomes more and more centralised and more problematic in the long end. So we try to create a consensus mechanism that allows a consensus protocol that agrees on what the network sees as quickly as the P2P layer can communicate these transactions. So all of a sudden you don't have this fixed time problem, you're just at the mercy of the topology of your network and you know we've done a bunch of stuff there as well to improve how quickly those messages can be spread by going to like boot nodes and then quickly being published out to a bunch of others. And then that's how we get a lot of our speed. So that was the one thing. And then we started building, so the last about year and a half, we've actually been building analytics tools to be able to see exactly what goes on in these systems because there's a lot of assumptions. So for right now, let me take a step back. So early, what was this? 2020, 2019. So we had finished our consensus engine. Now our consensus engine is really just an ordering engine, but we also knew we're going to need a virtual machine. And at the time we saw, hey EVM is the most adopted thing, so let's use the EVM. Exactly how it is out of the box, no modifications. So we took the EVM and we put it on top of our consensus agent. Cause the consensus it just does the transaction ordering and then you just push those transactions in that order through the VM so you can decouple sort of execution and consensus there. So we could plug in Ethereum and at one point we had a plan to add whatever else, but we've just stuck with the EVM. The Solidity and Viper I think have been the big winners in this ecosystem. Even though I think in this next way we're gonna see a transition away from the EVM, but not the languages. The languages are gonna stay the same. So if when we didn't have the EVM and there was just normal transactional throughput, we were hitting between 60,000 to 80,000 transactions per second in a sort of throttled network. When we added the EVM, that dropped down to 200. Consensus layer wasn't the problem at this point. EVM was the problem, but we weren't sure where is this problem, you know, is it and how the OP codes are executed, is it, and how the stack performs? Is it Merkel-Patricia Tree? Is it the underlying data store? We don't know. So we spent a last year of building these very detailed performance tools that can show us exactly on CPU, on disk, on memory, on time spent, on each thing that happens, where are these big bottlenecks? And then one of the big bottlenecks we found was storage. It wasn't CPU, it wasn't execution and it wasn't the upgrades. Those things are actually fine, it's the time it takes to write this information in the level DB or Badger DB or whatever DB implementation the specific node uses. We were using gif and gif users LevelDB, that was our benchmark. And something like I stand on correction, but I think it was like 98% of the time was spent just on DB reads and writes. So we started looking really into detail into these databases and now these databases, something like LevelDB, you know it's a standard generic database. So it's a standard relational database and it's not optimised for VM storage data, it's not optimised for hashes, it's not optimised for Merkel-Patricia trees. So it does a lot of things behind the scenes that's actually making it a lot slower. It builds up B3 indexes so you can do partial tech scans but that's not something you're ever going to do. You know a smart contract data structure. So you don't need that, you can throw that away. And so systematically we started going through these things and we originally tried to optimise the database and we moved away from LevelDB to PebbleDB and we did a bunch of optimisations there until we finally realised look we need a custom database system, we need a custom solution. So then for about the last 6-7 months we've been writing our own custom database that's specifically designed for hash lookups and it's specifically designed for key value stores based on smart contract data. So we take the smart contract address and the index of the storage slot and then that's your direct lookup. So now all of a sudden you don't have to try and traverse through a lengthy tree to try and get down to the Merkel route of where the thing is you are because you need to traverse up one leg every single time. You know exactly where it is because you know the smart contract into the storage slot and just that for example allowed us 98 point something percent less disk overhead because you don't have those index trees anymore, you don't have the big merkel tree, any of that stuff. So all of that stuff can be thrown away. So 98% less so where Ethereum was currently at 1.2 terabytes, with our new database that goes down to a hundred and something gigabytes and throughput wise in terms of read and writes, it's more than 8.4 times faster in terms of those lookups. So all of a sudden that 200 transactions per second now goes up to 4,000 to 8,000 transactions per second, which is already a huge change just because we looked at this one specific area. That's something we can do today that doesn't need to wait for data layers to increase or throughput to increase or any of that kind of stuff. But a big reason why we're focused on this stuff is because nodes currently are still too high requirement, still needs a server sitting somewhere in a data centre to actually run this stuff. We want a wallet and a validator to be the same thing. There should be no reason that those two are different. So if you have a wallet on your phone or in your browser, it should be able to be a validator and we're trying to get the tech stack that small that it can run, you know, inside of a chrome extension inside of a phone on a smartwatch, it shouldn't matter. And then be a validator. And we can get there, we know we can get there just on the tech stack alone. We already know our consensus is light enough to run on it. The big problem has been disk, you don't have 1.2 terabytes of storage to run in your phone. So obviously a hundred gigabytes is still too big, but we're getting closer and closer with each step we take. After this iteration, we run through a next loop and we improve the next thing until we get to the size we want. Because the last big bottleneck is the time to transmit between these networks. That's the slowest point currently. If you send a transaction in your browser wallet that goes to an RPC in Infura. That then goes to a node that sets somewhere in some server that then propagates to the other node so that they get it in their Mempool and then they can do a block and then it comes back and that whole process is going to be your slowest part because that you can't really decentralise at this point, not decentralised, distributed network at this point until you have so many touchpoints that you know, a point of sales system needs to be a validator, a wallet needs to be a validator. Anytime there's an interaction, if it is a validator, then all of a sudden the network layer doesn't matter anymore because in these small local environments you can quickly reach consensus, agree and then bubble it up into the rest of the network. And then we finally have something that can actually achieve our benchmark because now the other thing that's important is blockchains as a whole consider their benchmark Visa And Visa isn't our benchmark. Like blockchain isn't a payment network, blockchain is the decentralised web. Our benchmark is every single http request or TCP request that happens every second across the world. That's the benchmark we have to hit. Like we can already do Visa level throughput, that doesn't matter because we're not just Visa. This is like saying every traditional finance and payment systems runs on top of Visa and it doesn't. There's so much more we need from that. So all of these things I've been talking about earlier about this baseline that needs to be improved so that these new generation of apps can exist. Right now it's just financial apps but we're moving into a future where it's not just financial apps. It needs to be streaming, social media, gaming, whatever else it is. Again because it's the decentralised web, it's not just the decentralised finance ecosystem and there's a beautiful analog there again because if we compare it to early internet when you needed a data centre to have a single PC and you needed technicians that are capable of knowing how this thing works and you need special infrastructure to have a modem and you need to actually directly talk to an IP provider to give you an IP so you can actually host something there with reverse lookups. It was so hard to do and it was so expensive to do that. It was only used for two things, which was money or war. Because those are the only two that made sense. Fast forward 20 years and it became so cheap and so easy to access that now you access the Internet without knowing what a 56K dialup modem is or needing to get special hardware or technology and stuff. And it's so cheap that you don't even realise you're spending costs on it. But no one now that takes it all for granted could have done the stuff we're doing now back then. And it's exactly the same evolution we're going through with blockchain in that we are still sort of at that 56K dialup stage. We're starting to move to ISDN but we're still very far away from fibre. Very far. 2 to 3 years. We like we first have to get there for all of these other apps to be able to build. So that's Fantom's mission statement / value proposition and that's the base layer. And just to shill it a little bit more, the last what now the last three years have given me a fairly unique insight into sort of the struggles of dApp developers. So there's a lot of things that I didn't like that I was always like I wish the base layer allowed me to do this kind of stuff. So I've also brought a lot of that stuff to the table for Fantom. So examples of that includes gas monetisation, you know. So I hated the fact that if I launched my dApp, the only real way for me to make revenue is to add a token, because if you don't have a token and I add fees, someone's just gonna fork my code and they're just going to steal from my fees. So that's the only viable revenue model I had as a dApp developer. With Fantom, 15% of all of the gas spent on your contracts goes to the developer. So it doesn't go to the validator. The validator still gets the bulk of it and we'll probably still tweak those numbers. We just started small, we literally launched those 2 months ago. That gives them a revenue source. And the other reason I like that is because it grows as your user base grows, but it also allows the dApp to validate if they have a theory about something. Because now I don't have to go raise a bunch of money and then try and do my app. I can just taste it in a small scale and you know, if it works and it grows, I can always do that later. But I have some sustainable on chain funding that doesn't need me to go interact with a third party or anything like that. And then something else we did was gas subsidies. So the end user doesn't have to spend the gas fees, the smart contract itself can be subsidised. So whether that's why the dApp developer, the community, the foundation, someone can provide X amount of gas and provide rules within that gas is allowed to be spent, and then the end user that uses the dApp no longer needs to know about FTM. You know they can just use the app because that to me was a big onboarding problem because anytime somebody wanted to use my app so I was like okay first go get ETH, and then register in this exchange and then transfer from there and you have to do this and it's a tedious process and it makes it so much harder for you as a dApp developer to grow your app or your user base. And it's one thing with DeFi because people are financially incentivised to learn how to use this stuff, but if it's decentralised social media or streaming, users aren't gonna go through that effort to use your thing. So we have to be able to remove that. And then native smart wallets. So that allows the authentication to be deferred to something else. Whether that be username, password, face id, social auth, doesn't matter. You just need that. You don't need to go in and install meta mask and back up a seed phrase and try and store a private key. Because those things are scary. Again, they make sense why there's financial value but they don't make sense when it's a fun little gaming app or a social media app or stuff like that. So a lot of our focus is in what does that future blockchain look like, that is the decentralised web and how do we get there and how do we remove the pain points? And that's been our core focus. It's definitely evolved a lot from just started with, hey we have an idea how to improve consensus to where we are today. Writing our own databases and optimising our own stacks.
Walter Jennings: Well Andre, you've been credited with kind of solving the blockchain trilemma, which is balancing security, decentralisation and speed. And I feel that before too long when we're having much more on the chain, I would envisage a day where I would have my medical records, my passport, my tax files, et cetera, and you'll want to be able to get those immediately and know that they are extremely secure.
Andre Cronje: Correct. There's always a little bit of a trade off. There's no magical answer. We've definitely made some leaps, but even with a BFT, which is secure enough and it's definitely fast, but there is a little bit of a security trade off because currently, due to the hardware requirements, the validators need to be somewhat beefy. But again, as we decrease the storage and we decrease this memory and we decrease the CPU overhead, it becomes lighter and lighter and that means more and more participation. Anyway, long story short, we're getting there. I don't think we are there yet in terms of having solved the trilemma, but where we see how to get there and we understand the roadmap to get there and now it's just time and commitment to do the work to get across the finish line.
Walter Jennings: Andre, we have a segment here on our show called News in the Finoverse. Was wondering if there's any new developments that you personally want to introduce or Fantom want to introduce? What's the news in your corner of the world?
Andre Cronje: It's the same thing we've really been hammering on for a while, so it's not new, it's not sexy. Like the gas subsidies I think is really useful for dev. Gas monetisation I think is really useful for Dev. I think it's good for them to explore it, but at the same time we're pulling back on the throttle in terms of trying to get a lot of these dApp devs because we're very close to releasing this new database and we're very close to releasing a lot of these upgraded features. And when I say that, I mean Q3, Q4 this year and I'd rather have them come use it then, so that they get the full experience then come now it's six months. Which feels like a lot in blockchain terms, but it's actually, it's over in the blink of an eye.
Walter Jennings: Well I'm sure they'll be back with those upgrades.
Andre Cronje: Yeah, exactly. Just feels easier to do it then than to have them do it now and then potentially disrupt them later. So we've been kind of quiet on that front.
Walter Jennings: We'll uh, share a lot more than about the Fantom Developer Incentive program at that time. We don't want too many people rushing this stage.
Tracks in the Finoverse
Walter Jennings: Andre to close us out. We ask people a segment we call Tracks in the Finoverse. If there's music that powers your journey, gets you through the day, what would that be? And we'll play that as we're ending the segment today.
Andre Cronje: Yeah, look, I, I don't have music when I'm working on a difficult problem. Like even if I have music in the background, that's the first thing I pause. It's like when you're driving and you're listening to the radio and then you need to focus and get proper directions or something. The first thing you do is switch off the radio. It's very soothing when what you're doing is predetermined, but it's very distracting when you need full brain power. So I don't actually have music I currently listen to and when I do, I'm unfortunately very unoriginal. I literally go to YouTube and I click on the music tab and I click play the machines and the robots know what I like better than I do. You know what something I always liked, but this was many years ago, was Bittersweet Symphony by the Verve. That used to be one of my favourite songs. That should be nice.
Walter Jennings: That will be a great way to send us out because I'm afraid to show the listeners what would happen if we generated random music selection from my YouTube feed. It would be very different from yours. Andre Cronje, the director at Fantom Group, very well respected in the DeFi community. Thank you so much for joining Waves in the Finoverse.
Andre Cronje: I appreciate the time. Thanks.