About Solana: We are a lightning-fast distributed ledger technology for mission-critical decentralized apps. This podcast is a discussion between our staff, community developers, and industry leaders. You can follow us on Twitter @solana or GitHub @solana-labs. Subscribe on Spotify, Apple Podcasts, Google Podcasts, or the direct rss feed.
Here is the show transcription:
Andrew: Hello everybody, and welcome to No Sharding, which is the first episode of the Solana Podcast. I’m here with Greg Fitzgerald. Hey, Greg.
Greg Fitzgerald: Hey, Andrew.
Andrew: And I’m Andrew Hyde, and we’re going to talk a lot about Solana on this podcast over the next couple weeks and months and years, but the first thing I want to say is just let’s do an introduction of what Solana is. So Greg, you’re one of the founders of Solana. What is is Solana?
Greg Fitzgerald: Solana’s a very high performance blockchain. Typically, a popular blockchain like Bitcoin and Ethereum could only do about 10 to 15 transactions per second worldwide, and Solana’s aiming to do hundreds of thousands of transactions per second worldwide, enough to be like a NASDAQ of the blockchain world.
Andrew: I love it, and I think that one big differentiator of Solana is that we’re actually saying it’s the same as Ethereum. It’s decentralized, it’s secure. We’re creating a protocol that actually meets all those standards yet can scale.
Greg Fitzgerald: That’s right.
Andrew: So the industry has been an amazing industry to watch, but it’s gotten to this point where everybody’s just kind of building to the point and then saying, “Oh, my goodness, this is too slow. I can’t build on it anymore.”
Greg Fitzgerald: Yeah, it sounds like it’s super-frustrating for the Ethereum communities especially to be able to create these really interesting programs on blockchain that could do just about anything but not do anything very well.
Andrew: Yeah, one of my favorite members of the community described blockchain as a slow database that’s great for groups of people that don’t trust each other.
Greg Fitzgerald: Okay.
Andrew: And I thought that was a great definition until I met Solana, right? Because Solana’s an incredibly quick database. Or not database, but you can look at it like a database. I’ve been describing it that way.
Greg Fitzgerald: Yeah, absolutely. Right when we started the project, we really thought about, “Okay, it’s like a database,” and so rather than starting from scratch with performance, we said, “Why doesn’t this thing perform anywhere in the ballpark of a centralized database?” So we started with that as the bar, and said, “Okay, if we add these decentralization, permission-less, these great, amazing properties of blockchain, how much slower does it have to be?” We really found out it’s not much.
Andrew: Which is amazing. And for people listening that have been around cryptocurrency for a while, that doesn’t sound novel. But if you actually are in the weeds in crypto and understand the tech, you understand how difficult it is to build a decentralized system that actually is trusted, fast, and meets all the boxes of what we believe in.
Greg: Yeah, and it’s perfectly understandable in my opinion. If you look at the history of these guys coming in from the cryptography space, where performance and making a system that scales to everyone in the world was not a priority. They were proving out the cryptography, that this works, that you can have key pairs be used to determine who owns an asset, and how to transfer that asset. And they did that. They proved that out. But the way they proved that out, and when they kind of add on these features that we’re more familiar with for general purpose programming language, that’s not their area of expertise and it really shows.
Andrew: So it’s a good, I mean it’s a really good, first attempt, and it totally took the world by storm. We love it. Now what’s going to happen next over the next five years? So let’s look at Solana, and in turn tell me about the founding team.
Greg: All right. The naming Solana, it’s named after a beach town, Solana Beach, just north of San Diego, where Anatoly and Stephen, the two other co-founders and they used to live, we used to surf there every day. We just had a lot of amazing memories of that place, and we named the company that to … A good old nod to Solana Beach.
Andrew: That’s awesome. So you’ve been together a while.
Greg Fitzgerald: Yeah. Oh, man. 15 years, I think we’ve been writing software together now.
Andrew: And where did you write software together?
Greg Fitzgerald: At Qualcomm. We were in … Actually a lot of the core Solana team is from the original Brew team. They were a mobile operating system that Qualcomm developed … Before there was Android and iOS, there was Blackberry and Brew. You may not have heard of Brew so much because we pretty much just stamped whoever he carrier was on top of the Brew operating system, but I believe we were the market … Sorry, what’s the-
Andrew: Market leader?
Greg: Market leader, sorry. Can you tell that I’m not in the marketing?
Andrew: Yeah. So I’m with marketing. You’re the CTO. So I think that there’s going to be some things that we say that really connect with more of our CTO friends, and there’s some things where we say that I’m going to try to connect with the marketing folks. So it’s a good tag team on that.
Andrew: So, you built market-leading software that everybody in the world knew of, if used.
Greg Fitzgerald: Yeah, we knew smaller, faster, cheaper, better than anybody else in the world, and now we’re applying those concepts to blockchain wherein nobody else has yet.
Andrew: When I talk to people that’ve read the white paper, because the Solana white paper is fantastic. If you read the Solana docs … We can link to that in the podcast notes. But this is written by people that have been working on assembly code level computer programming for a long time. This is not, “We’re going to fork a programming language and try to make it 10% faster.”
Greg Fitzgerald: Yeah, apologies for that.
Andrew: Well, I don’t think there’s apologies for that. If you’re trying to build the fastest solution in the world, which can we say 50,000 transactions a second?
Greg Fitzgerald: Oh yeah, definitely, comfortably.
Andrew: Yeah. So we’re comfortably in that. You don’t want to speculate on what you can do because it’s unproven. But 50,000 transactions a second is what we’re seeing right now on our tests now.
Greg Fitzgerald: Yeah.
Andrew: That is only achieved by a team that’s worked together at the base level for a long time, in my opinion.
Greg Fitzgerald: It’s helped a whole lot, that’s for sure.
Andrew: So tell me about Anatoly.
Greg Fitzgerald: Anatoly, he was the operating systems guy. I was the languages and compiler guy. Our buddy Stephen was the GPU expert. And what Anatoly really knew well … You know on the distributed systems protocols, especially like to apply those to operating systems to get normal operating systems closer to distributed operating systems, had that perspective.
Greg Fitzgerald: And the thing that we just kept going back to, and a lot of the stuff he invented at Qualcomm there, was about how to find a way to get different worlds to communicate to each other efficiently. So in Qualcomm, it was getting the ARM processor to talk to the Hexagon processor. Then at Dropbox it was about getting the nodes that are supporting the file system to talk to each other. And now in blockchain, it’s about how to get these nodes, these validator nodes that don’t trust each other at all, how to get those things to talk to each other, which is considerably harder but always seems to coming back to the same solution, which is how do you get these things to synchronize? How do you get a clock into this system?
Greg Fitzgerald: And in blockchains, that’s actually a lot harder. That’s turned out to require some real innovation. That’s where Anatoly came to me. He says, “Hey, I got an idea how to get a clock into blockchain.”
Andrew: Because nobody’s done that before to my knowledge.
Greg Fitzgerald: Yeah. Well, right. There’s been kind of two techniques used. You can either use asynchronous protocols that don’t require a clock, that just requires a lot of messages, sometimes to the order of N cubed. But and then the other solution is, and Ethereum went this route, is to have this notion of a median timestamp, where you have all these notes that don’t trust each other go submit their notion of time, and they just sort of guess which one of those is close enough to being true time and they just run with it.
Greg Fitzgerald: And that is very clunky, very slow, and the asynchronous ones require a lot of messages. And so, Anatoly came up with this brilliant mechanism which we called “proof of history,” where you can have a very, very fast clock in an environment where the nodes don’t trust each other at all. And by having that, we’re able to use, we’re able to leverage these synchronous distributed systems protocols, which are must simpler and much faster.
Andrew: So if you’re listening to this, I want you to just take one thing away from this conversation, which is, Solana equals proof of history, right? Or the big innovation that Solana’s bringing, and trying to get the world to adopt, I would argue, is proof of history.
Greg Fitzgerald: Yeah.
Andrew: Can you explain to me, like let’s just go very basic, what is proof of history?
Greg Fitzgerald: It’s a notion of time before consensus. And what is that? If you know bitcoin and its proof of work, bitcoin actually implements something remarkably similar to proof of history where you do this proof of work, you solve this puzzle that take approximately 10 minutes to solve, and that’s when everyone generates a new block. So actually, bitcoin has this trustless mechanism to create a tick on a clock every 10 minutes. Well, that’s great, and you can actually build things on it. You can postdate a transaction to being hours ahead of time. But that’s all after the consensus mechanism. You’re talking units of 10 minutes.
Greg Fitzgerald: We need stuff in the order milliseconds, microseconds, to build a consensus on, to build sub-second fidelity on. And so, we used actually that same mechanism, a proof of work, but shrunk it down to actually nanoseconds for each unit. We stacked those up to create this sort of water clock. You imagine a clock just dripping drops of water, and however many drops, that’s kind of a clunky notion of time, a clunky notion of duration.
Greg Fitzgerald: But even though it’s clunky, you can say that a certain amount of time definitely passed. There’s water in the bottom area, right? And so we do that same mechanism. We have a very fast, very clunky mechanism of duration that is completely trustworthy from a cryptographic perspective, and that’s what we call proof of history.
Andrew: I like that a lot. If I were to describe a bitcoin transaction to somebody, I would say, it’s person A sends something to person B, and there might be 100 computers along the way that do little, tiny math problems, and if everything fits in perfectly, at the end, person B, the transaction goes through.
Andrew: And people, some of those computers, that did math problems along the way get rewarded. That’s the very, very basic.
Greg Fitzgerald: Yep.
Andrew: With proof of history, it’s almost like we’re getting computers to argue over what time it is, and that’s actually significantly secure. Is that right?
Greg Fitzgerald: That’s right, yeah. The time moves forward as quickly as each computer can, and the one that has demonstrated that it has done a certain amount of work the fastest is the one that basically gets to say, this is what time it is, and everybody else needs to find a way to do this work a little bit faster.
Greg Fitzgerald: But it’s not competitive in the way proof of work is, where you need more and more GPUs, and eventually you’re using the same power as the country of Iceland. This is just competitive on a single CPU to make the most of what you can on that one processor.
Andrew: So it’s extremely efficient. We’re doing sub-second finality, which is unheard of in the industry, in the decentralized blockchain industry. And then we’re also doing it extremely efficiently, so we’re seeing 1/1,000th of the power usage. [crosstalk 00:13:19] blockchain, while being 100,000 times-ish faster.
Greg Fitzgerald: Yes. That sums it up nicely.
Andrew: So usually, when I get to this point of a conversation with somebody, they say something to the effect of, “Yeah right, that’s not true. You can’t do that.”
Greg Fitzgerald: Yeah, we get that a lot.
Andrew: We have so many people in this industry, so many smart people working together, and we’re going to roll out the next little bit that’s going to make it 10% faster, but nobody solved the problem this way. So the founding team, having worked on this for the last 15 years, off and on with their expertise, seemingly has done this. But what do you say to people who are like, “I don’t believe it?”
Greg Fitzgerald: Yeah, it’s a hard pill to swallow. I understand that completely. The way I like to describe it is, it’s really a different set of expertise coming into this space, where the cryptographers have been iterating regularly with the expertise that’s in close a proximity to them. We have come in with embedded systems expertise, and that brings in a very different perspective of how to go about doing things, and the way we’re doing things is just a lot closer to the way centralized databases are built as well, and that’s really what was needed in the blockchain community, and just wasn’t there, I think.
Andrew: This analogy might fall flat on its face, but it’s almost like we have combustion engines powering cars, and then suddenly have electric cars, right? And everybody’s on the racetrack, going like, “Yeah, I’m going zero to 60 in this many seconds,” and the electric cars are coming out just like, “No, no, no, we know how to be extremely efficiently at torque. We know how to …” Yeah.
Greg Fitzgerald: Yeah, and there is a lot of the expertise that we have, there’s plenty of that around the world. And I think we’ll credibility that people will believe what they’re seeing as we see more and more people come in and they’re able to hit these same numbers, and hopefully we’re on to bigger and better things by then and people can look back and say, “Oh yeah, they actually did that. Not a scam.”
Andrew: Yeah. I mean, it’s open-source software, right?
Greg Fitzgerald: Uh-huh (affirmative).
Andrew: Right? So I can go to GitHub right now, look up Solana. I think it’s Solana Labs.
Greg Fitzgerald: Mm-hmm (affirmative).
Andrew: … And check out every bit of code you’ve written.
Greg Fitzgerald: Yep, yep. You can see it for yourself. Software’s difficult. That’s not always sufficient of a convincing argument. There are magic tricks that can be done in this thing, shortcuts that can be taken, so to those people I say, not only do that, but go get your smart friend that you trust and have them take a look at it and convince you that, yeah, this is real.
Andrew: Cool. So let’s talk about how to get involved or how to actually view that this is real or live. Let’s say live, not real, because every person I’ve talked to … I was in San Francisco watching a live code review happen with some founders that had built a blockchain, and they were … I’m going to label you as extremely smart if you launch your own chain. If you’ve written your own thing, I’m going to label you as extremely smart.
Andrew: And Anatoly was explaining everything, and like, “This is how we do leaders, and this is how we do,” a couple of other things, and there was this silence. And they just stopped and said, “You’re very smart.” And they were kind of annoyed at how clever this design was.
Andrew: So if we can, let’s talk about the book. So on GitHub, and we’ll link this in the show notes, there’s kind of like the intro book. Can you tell me about that?
Greg Fitzgerald: Yep. I wrote that myself, actually, and it’s kind of the second iteration. So in the first iteration, Anatoly wrote the white paper that’s mostly about proof of history. We built that up and then found out that there’s actually a whole bunch of other performance bottlenecks that affect blockchain. And we went and solved all of those problems. Pipelining, storage, account storage as well, not just ledger storage, all of these different problems that affect performance. So we built out all of that kind of in isolation without updating the white paper at that time. And then once we had all that stuff and were able to demonstrate the full blockchain operating at the kind of speeds that we had anticipated were possible, we rewrote that whole book.
Greg Fitzgerald: And so now, I think that’s probably the best spot to go right now and get a good, high level overview of all the different technologies that we’ve stitched together here to make this thing work.
Andrew: And you can read just a short intro, a short explanation of what the project is, and then you can actually play around with some code. So let’s talk about the testnet and what would it look like for somebody to actually start hammering at the code?
Greg Fitzgerald: Let’s see. So you’d start with that book, and in that book, there’s a getting started guide that tells you the commands to executive on the command line to sync the code and to build it. And it’s really easy. Actually, we’ve got scripts in there that take care of all of the dependencies. It’s all automated. And then with just a few lines of code there that you type in, you can start your own local testnet.
Greg Fitzgerald: And then right after that there’s another in that book that says how you can download this Tic-Tac-Toe app that’ll interact with either your local blockchain, or you can interact with the testnet with that same code just by retargeting the URL.
Andrew: And if you want to see it live, you can actually play Tic-Tac-Toe right now.
Greg Fitzgerald: Oh, yeah, yeah.
Andrew: So if you go to the GitHub book or you go to blog.solana.com/developers. You can find a link to two really interesting projects. One of them’s Tic-Tac-Toe where you can play Tic-Tac-Toe with somebody on blockchain. Everything’s recorded, and they can see all the hashes and what it actually looks like.
Andrew: And the other one is messaging. So you have messaging without a sign in, that is decentralized, that is quick, that is … Do you want to anything more about the messaging?
Greg Fitzgerald: It’s just the most amazing thing about blockchain that you can just create a form of identity on the fly, that this address space is so massive, big enough that there’s one for every atom in the universe, right, that you can just randomly point in a direction and say, “This one’s mine. I’m going to call this one Greg.” Yeah, it can do that. Of course, just for playing a game of Tic-Tac-Toe or having messages, but for anything. It’s a completely permission-less way to say, “I am going to take this identity and I’m going to associate it with this asset, and if I can create a reputation that this identity is good for exchanging that asset, then I can do it. I can create an exchange without anyone’s permission, just get up and do it.”
Andrew: For the non-CTO crowd, that means that your browser can be your identity. That means your phone. That means you don’t have to log in with some things to actually be able to interact with things.
Greg Fitzgerald: You don’t have to go to a carrier and ask them to register, you just do it. You just say, “This is me.”
Andrew: And then nobody owns that, per se.
Andrew: Right, like the community kind of owns that.
Greg Fitzgerald: Mm-hmm (affirmative).
Andrew: And we’ll dig into that, those concepts, in future podcast episodes. And I think that there’s a lot of style involved with that. The people, they see the world a certain way. They see computers and how society interacts with technology in a certain way, and I think that this empowers people to build projects, and actually build them, and not just theoreticize about them. And actually, let’s actually build the platform, or let’s build a mapping platform, or let’s build an alternative to Facebook …
Andrew: There’s all these dreams that these people have, but then you’re kind of bottlenecked on the speed of the network.
Greg Fitzgerald: Mm-hmm (affirmative).
Andrew: Or the cost of the network. So, with this incredible speed, Solana’s seemingly opened the floodgates.
Greg Fitzgerald: Yeah, just from the cost standpoint alone, you’d typically have transactions at a normal financial system costing 10 cents plus a percentage on top of that. We can do millions of these transactions for a fraction of a cent. And because there’s not thousands of humans in the middle, there’s not buildings full of people working to make these things happen, it’s all software, and the integrity of this ledger is maintained with cryptography, with software, so it’s only as expensive as the power to run these computers.
Andrew: That breaks my mind to think about, a global supercomputer that you can tap into and write applications on.
Greg: Yeah, it is outside of any borders and will just keep going so long as anyone in any country decides to put one of these validator nodes on the network and say, “I’ll help maintain this ledger.”
Andrew: All right, so let’s dig into validators and nods. Explain it to my marketing mind. Explain what a validator, what a node is.
Greg Fitzgerald: Well, validators are basically sort of the trick to not require a centralized organization to be trusted to maintain the integrity of the transactions. Instead, anybody, anywhere, can say that this person wanted to transact this many tokens, we say, this currency, to another, and it becomes these validators’ job to actually agree on whether a person actually did want to do that with the tokens they supposedly are associated with their private key.
Greg Fitzgerald: And so the validators, they all follow this protocol, this software, where they can verify the client’s intent, and the validators can choose which of those transactions actually occurred, batch as many as the possibly can to ensure that as many go through, and that this currency system actually works.
Greg Fitzgerald: And yeah, that’s kind of the trick, is getting these things to come to consensus on all the stuff that users were trying to do. And it solves the problem of a user saying that they want to spend their own tokens more than once, so it actually does have … At some point, the record of the transaction, it does have to come to some centralized spot. They do have to agree that if you want to spend somebody some tokens that you actually haven’t already sent them to somebody else.
Andrew: The double-spend problem.
Greg Fitzgerald: The double-spend problem, right.
Andrew: So our validators, and we just launched actually about an hour ago … We launched Tour De Sol, which is the Solana testnet event. So we’re challenging people to break out their Linux boxes or their interest in Linux and if you’re on a gigabit … Actually, there’s no technical limitations to having a validator on the Solana network.
Greg Fitzgerald: Not right now, not for the first stage. I think it doesn’t take much. We’ll do more advanced things of getting high throughput use cases a little later down the line where we’ll need GPU cards involved.
Andrew: And you can go to blog.solana.com/TDS, Tour De Sol to learn more about the testnet event. But Solana’s got a pretty big road biking culture. Can we talk about that? I mean, we’re doing Tour De as our event name.
Greg Fitzgerald: Yeah. Yeah. Let’s see … Of the founders, four … All five of us, actually, ride a bunch. Four of us have done Iron Man Triathlon. So yeah, we’re pretty into it.
Andrew: I mean, I’ve ridden my bike across the country. There’s something about culture around companies that I find very, very interesting, where you find some people that just latch onto certain sports.
Greg Fitzgerald: Yeah, and it’s kind of fun to see in a sport where everyone’s so focused on performance, going faster a little, it definitely does carry over into the software, the competitive nature, the being satisfied with just going a little faster every day.
Andrew: And the beauty of the machine.
Greg Fitzgerald: Uh-huh (affirmative). Sure.
Andrew: A bicycle’s a beautiful machine.
Greg Fitzgerald: Yeah, we like beautiful machines.
Andrew: And we’re here building beautiful software. That’s really cool. What’s the other sport? What’s the other Solana sport?
Greg Fitzgerald: Oh, you dropped it. Now you got to explain it. Yep, it’s underwater hockey.
Andrew: Underwater hockey, which is not a joke.
Greg Fitzgerald: Yeah. Only three of the five founders play that one.
Andrew: And you play it as an elite level as well.
Greg Fitzgerald: Yeah.
Andrew: You were just at the national championships. You’ve been to the world championships.
Greg Fitzgerald: Yep, got a bronze medal this year at my old age, so I’m feeling pretty good about that. It’s good to be back in that position again after a lot of years. Used to play at the elite level a long time ago.
Andrew: It’s one of the most hardcore sports I’ve ever seen as far as endurance, intelligence, brute force, brute strategy.
Greg Fitzgerald: Yeah.
Andrew: It’s a really interesting sport.
Greg Fitzgerald: Four-dimensional sport. You go up, down, left, right, and you’re holding your breath, and trying to figure out when you’re going to run out of air, and what you can do with that time you got.
Andrew: Sounds simple. Sounds just like a startup.
Andrew: So Solana is … All rounds lead to Main Net. I’ve said that a couple times on blog posts and such, but you’re a project that has so many employees now? I’m going to say over 15, under 20.
Greg: Mm-hmm (affirmative). 15 to 20. Seems to change every day at this point.
Andrew: Yeah. Explosive growth lately. I mean really, we’re looking at that … I think you guys went … And I say “you guys,” as far as I’m a new member to the team. I’ve been here for a month now. But you’ve been heads down for a while without really marketing. Do you want to tell me about that and what that choice … Or why you chose to do it that way?
Greg: Well, it actually has a lot to do with crypto winter, really. I think the community thinks we’re looking really good in the world when we first raised money to build the project, and then the sentiment got very bad. People thought that maybe this blockchain thing wasn’t going to happen. But we’re true believers, that’s for sure, and we decided to take that as an incredible opportunity, really, that we didn’t need to spend any time or money or putting out any notion, or being concerned of any notion of vaporware. Instead, we could just build this stuff, make it real, and when winter is over, we can come out and make just one hell of a splash with a product that works that with can deliver on everything we promised.
Andrew: It’s a really interesting thing when people hear about it, because it’s really … As far as tech announcements go, this is huge, right? 15,000 transactions a second is absolutely massive for the industry, for it to be truly decentralized. And a lot of people have basically spun up Amazon Web Services accounts and said, “We’re super fast!” Well, it’s like, you’re centralized. You’re a database. Let’s actually build a blockchain that’s super, super fast.
Andrew: But you did it, and you did it in a way where you didn’t really talk too much about it. You kind of went into hiding for a year. I don’t think you’d describe it as that, but I think the industry [crosstalk 00:29:50].
Greg: No, we totally went into hiding. There’s no video. I didn’t get a haircut for like a year.
Andrew: We’ll be taking a photo and putting it on the share nodes. But we shared an office a year and a half ago, and then I didn’t see you for a while, I didn’t know how the project was doing at all. And then suddenly, the testnet’s live.
Greg: It was nice to have the opportunity to go that deep, to really just focus on building great code that worked well and not have to explain every step along the way, or promises, and software is hard. Software doesn’t always go to plan. It didn’t for us, either. We had a lot of hurdles along the way, things we thought were going to be ready to go, and then it turned out, didn’t quite work as we hoped and so we had to go back to the drawing board. And it was really nice to be able to do that all without disappointing the public in that we wouldn’t be able to meet that anticipated roadmap.
Andrew: So you kind of had no pressure, and you kind of went … And I think about it-
Greg: I don’t think anyone that works for me would say that there’s no pressure.
Andrew: Fair. But you got to go and write and have that luxury, where a lot of projects will launch and then have this massive pressure, make a lot of shortcuts based on that pressure, and then not be proud of the product they launched with. And it sounds like you’ve had the luxury of just being able to write that, and the fortunate ability to have that team to back you up, to really build on that.
Andrew: I think I listened to Anatoly on a podcast. He said they started sleeping about two months ago when testnet went live.
Greg: That sounds, yeah, that sounds just about right.
Andrew: And suddenly he also had a child, so I bet there’s a lot of not sleeping with that. But testnet … And you can see there’s some really nice graphs if you go into the developer resources. There’s some really nice … This testnet is live. There’s validators in six continents around the world. It’s blazing fast speed. It’s not operating on full capacity right now, because A, testnet, and B, early on. But that’s going to start ramping up. So Tour De Sol, you’re going to start seeing more and more validators. Hopefully somebody listening to this is going to set up one and play around with that code, and I think if you have a good Internet connection and a nice Linux box, you’re going to be pretty interested in running the Solana Suite, the Solana validator codes … I’m not sure how to say this.
Andrew: So where do you see the project in, let’s say, a year?
Greg: The project in a year … I think just a year is so soon. Yeah, we’ll be launched at that point, and we’ll be really, I think, supporting the developer a lot, really looking at as Libra comes up to speed here, to see if their move language gains traction at all with the developer community. It’s currently not very pleasant to work with. It actually requires a lot to learn, so they’re Facebook, so they might just force the development community to learn some things there. So that’ll be a really interesting development.
Greg: As it turns out, our runtime is very, very well-suited to execute that language, so if that does go well, I think we can do what they’re trying to do better than they’re trying to do it, so that should be a fun development to see.
Andrew: And it’s fun to see Facebook step up with a big, ambitious product.
Greg: Yeah, oh, absolutely.
Andrew: And I think for us, we’re seeing 2.6 billion wallets being created. We’re seeing people have the ability to use crypto in a way that they haven’t in the past.
Greg: Yeah, it’s been really validating to read through their white papers, and they’ve got a lot of brilliant engineers working on that. I worked with one briefly on the LVM compiler tool chain on the ARM backend. It was always very impressive, his work, and he’s the head of blockchain over there. And reading through the white paper and seeing what they came to, a lot of the same conclusions that we did in how to do this, that there’s really only that one rather massive turn at the end there, to develop the applications on the system with either a new programming language, is what they did, and we chose to offer more of an operating system that will allow developers to code in any language that can target LVM and therefore target this BPF byte code for safe execution.
Greg: So our programming environment is a little awkward in some ways. Theirs is a little awkward in different ways, and I just can’t wait to see how it plays out, and see which form of the awkward the developer community might tolerate.
Andrew: I think that we’re going to start wrapping up with this interview. So great to chat with you and learn much more about the project. I think if I were to personally take a look at crypto and what I love and don’t love about it … I love the actual uses of it when people are actually using it. I’m not a huge fan of speculation around it, like the whole price speculation, let’s get rich by investing early. I don’t love it all. But I love the people, the developer community, that really wants to use this.
Andrew: And so I’m so in love with Solana, and I decided to join in part because I just see the developer community loving this. I think it’s such an interesting project.
Greg: Yeah, me too. I can’t wait to see what all … The kind of applications that come out of this. I think what we saw with Ethereum is just like, hey, this existing application that we’re familiar with cannot work on blockchain. Can we get any adoption? Maybe we’ll slowly build up from there. But in this world, where we have all this capacity, and this massively cheap transactions and high throughput, what type of applications are people going to come up with that basically just couldn’t be done on top of Visa, that the transaction price is too high?
Greg: But now it’s possible. I think we’re going to see the success stories are going to be immediately global success stories. They’re not going to be this year after year of kind of working their way towards adoption. It’s either going to happen or it’s not, and I just can’t wait to see.
Andrew: So excited, too. In a year, maybe we could do an interview where we just say … Maybe our next interview is, “Where do we see it in a year?” And then we wait a year and interview, look back.
Andrew: I think let’s end this with, let’s talk about the speed of Solana.
Greg: Yeah? What about?
Andrew: So what’s the fastest database in the world right now? Production-wise, the most-used, fastest database? I think it’s Visa or NASDAQ.
Greg: Well, NASDAQ is considerably faster. They’ll do about 500,000 transactions per second, and Visa peaks, I think, meeting 65,000 transactions per second.
Andrew: And in the white paper, if Solana works, and this is theoretical, where’s Solana?
Greg: We’re talking a theoretical capacity of 710,000 transactions per second while running on a gigabit network. If we put the same architecture on a 10 gigabit network, we expect that we can scale up to 10 times that, seven million transactions per second. On a 40 gigabit network, 28.4 million transactions per second.
Andrew: In theory, that hasn’t been tested.
Greg: Right, right.
Andrew: But correctly if I’m wrong here. That means that at the current one gigabit speed, Solana could power the NASDAQ and Visa network and every blockchain project out there at the same time and not reach capacity.
Greg: Yep, and by the time we get all that together, we’ll be out working on the 10 gigabit solution, be able to take it to the whole next level.
Andrew: Some projects that have been building and developers out there that have been doing some on-chain and some off-chain, we’re just inviting them to say-
Greg: Just put it all on chain. It’s too cheap to mess with that, to mess with trying to have your software work on multiple places and securing it in multiple places. Ours, it’s just so fast, so inexpensive, you can just put all the data right on chain.
Andrew: That’s amazing. So you can reach out to us. The best way to contact us is on Discord. So go to blog.solana.com, and upper right-hand corner, there’s a chat with us on Discord, which is a nice chat program. Twitter, SolanaLabs. What’s your Twitter?
Greg: Hell if I know.
Andrew: You’re not a big Twitter person.
Greg: I make the bits move faster.
Andrew: Okay, so I’ll keep the marketing. You can follow me, @unicorn. And then email … Do you like email to be out there, or should I just throw out mine?
Greg: All yours, Andrew. I’ll keep coding.
Andrew: If you have any questions, comments, about the name as well. I mean, I’m naming this podcast No Sharding, because if you listen closely, we never talk about sharding. We’ve got this insanely fast protocol, but not using the industry’s favorite term, sharding.
Greg: Yep, sharding’s easy to start, very, very difficult to finish. It adds a whole lot of complexity to the system, and we see it being able to work within increasing bandwidth year after year without ever needing to take on that complexity … That has a very strange choice that a lot of the other blockchains are choosing from the perspective of distributed systems theory, and we’re not going to go that way.
Andrew: So, no sharding.
Greg: No sharding.
Andrew: I love it. That’s a great way to end it. Thank you so much, Greg. That was the first episode, and I’m just excited for this project.
Greg Fitzgerald: Well, thanks, Andrew. I’m looking forward to seeing what’s next.
Andrew: This is going to be an interesting year, for sure.