Serum Initration - The Last Hackathon I Won

Tommy and Taylor are the founders of PsyOptions, a DAO developing the leading options primitives on Solana.

00:09 - Introduction and Origin Story

04:04 - What are the challenges / improvements in Solana?

11:49 - Integration of Serum v3

14:32 - ​​Adoption of PsyOptions

17:19 - Architecting the system

22:11 - Liquidity mining vs. options trading

26:27 - Background in trading options

28:05 - DeFi vs. Traditional finance products

30:56 - Gaming as a market

32:56 - Exciting things out of the hackathon

34:09 - Announcements for PsyOptions

37:53 - If Solana could change one thing?

DISCLAIMER

The information on this podcast is provided for educational, informational, and entertainment purposes only, without any express or implied warranty of any kind, including warranties of accuracy, completeness, or fitness for any particular purpose.

The information contained in or provided from or through this podcast is not intended to be and does not constitute financial advice, investment advice, trading advice, or any other advice.

The information on this podcast is general in nature and is not specific to you, the user or anyone else. You should not make any decision, financial, investment, trading or otherwise, based on any of the information presented on this podcast without undertaking independent due diligence and consultation with a professional broker or financial advisor.

Anatoly (00:09):

Hey folks, this is Anatoly and you're listening to The Solana Podcast. And today I have with me Tommy and Taylor, co-founders of the PsyOptions protocol. Awesome to have you guys.

Tommy (00:18):

Thanks for having us.

Taylor (00:19):

Thanks for having us.

Anatoly (00:21):

Cool. So what's the origin story? How did you guys get into crypto and what made you build PsyOptions?

Tommy (00:27):

Crypto, it goes back to... I remember watching the Ethereum ICO, just being a broke college student, but felt we were too broke to actually throw anything into and that's a big regret, but that shaped up how we got into Solana later on. Really dove deep into everything back in 2017, right before the summer hype. And then in the summer hype, tried developing a little bit on Ethereum, doing some solidity development in the spare time, but I never jumped full time into it until PsyOptions. Taylor has a little bit of a different history with crypto.

Taylor (01:03):

Yeah. I've actually been full-time in crypto since late 2017, after Tommy and I shut down a previous business we started in school. We were looking for different things to do and I knew crypto had a lot of hype in 2017. I was like, "All right, this is definitely an industry I could see myself being a part of." I eventually took a job at Blockfolio and then as well as doing some freelance solidity development and then been full-time ever since.

Anatoly (01:27):

How did you guys meet? What was the genesis for you guys to go build PsyOptions?

Tommy (01:32):

Well, Taylor and I are twins, so we met a long, long time ago. We've always been hacking on ideas and stuff. And I guess, Taylor had his eye on Solana from 2018, right Taylor?

Taylor (01:47):

Yeah, pretty early on. I remember Multicoin writing about it. I was like, "Oh, this is actually a really sweet architecture, solves a lot of problems that we saw in Ethereum." And kept following before Mainnet beta was launched.

Tommy (01:59):

Yeah. And so we had been tinkering around, created a GitHub organization last summer, like the same one we're using now and just started reading the documentation. And then had a few projects we tried in the fall that never really took off. And then in October we were surfing with Tristan from FTX and he was just talking about Serum and everything that they were working on. So we knew what was in the pipeline and had that in the back of our mind. We did the first hackathon, did in place, built a trusted third party Oracle. And then after that had an issue with TradFi, trying to get API access to automate a options trading strategy, and that was what kicked it off. We were for fresh off that first hackathon, wanted a fresh idea, had our feet wet in Solana. And it was like, "Taylor, what if we just built options into the blockchain? We can get this API access built in. We have the order book already there, there's some basic infrastructure." And that was the genesis.

Anatoly (02:59):

That's awesome. Limited access to data was one of the reasons I started building this thing. Because I used to try to build stupid deep learning models on interactive brokers and you never have access to data. It's always even the quality is really suspects. It's like, "Do I really know that this is where things got executed? Or did they just copy and paste stuff from a database with a bunch of errors?"

Tommy (03:26):

Yep.

Taylor (03:26):

That's terrible. Yeah. If you want good data quality, you have to pay up for it. That's why Bloomberg Terminal is what 20, 25K a month. And if you're just the hobbyist or just trying stuff out, it's just not feasible to pay that much.

Anatoly (03:41):

Yeah. This is to me I think part of the beauty of the space right now, is that you can build up a lot of what finance is with just a bunch of hobbyist. It's like Linux. Linux in the '90s, you're competing with Microsoft, billions of dollars of engineers buildings stuff, but it's just a bunch of people over the weekend can compete. It's crazy.

Tommy (04:03):

Yeah. It's wild.

Taylor (04:04):

I think that's one of the best parts, all that coordination.

Anatoly (04:07):

So what are the challenges? You guys are one of the earliest I would say teams working on Solana. What have you guys seen, or what were the real painful points? What got better? What still sucks?

Tommy (04:18):

Oh man. All right. This first Solana season hackathon, the one that we won, we wrote everything in Solana native. I remember pinging [Armani 00:04:29] back in February saying, "Hey, I hear you're working on some framework, can I poke around? And checking out the repository." But it wasn't anywhere near complete or, I didn't dive in enough to use it for the hackathon. So now I rewrote the entire American option protocol in Anchor and it took me very little time to actually write that. So the development life cycle and just ease of getting up to speed, has improved ridiculously.

Taylor (05:07):

Yeah. And documentation has improved too.

Anatoly (05:07):

That's awesome. What is Anchor doing for you guys that Native, Rust isn't?

Taylor (05:13):

It's helped simplify our integration tests. So that's one thing that we try to do when we first started was, we wrote our own integration testing framework in Rust. I guess I wouldn't even call it a framework, it was pretty rough. But Anchor takes care of that. You're just writing your test in JavaScript, it's pretty easy to get up and running. And then also handling a lot of different edge cases that you wouldn't have to think about, checking account addresses and other things just to bring safety in. And it removes a lot of those headaches that, if you're just getting started and trying to hack something together, you're not really going to be thinking about.

Tommy (05:48):

Yeah. I think the account, de-serializing accounts, token accounts and things like that. You just have your accounts structure, passing that into the context and it de-serializes all that. The amount of little issues we had just because oh, we mis-ordered one thing in the array when we were refactoring, the accounts array, and it's like, "What the hell is going on?" And then you're trying to debug and add messages and stuff, because you're just like, "Oh man. And what is..." And then it all turns out to be a typo or you fat fingered moving one line up or, and it was... So the account structure and dealing with that is just incredibly easy. You don't have to de-serialize anything yourself.

Well, anything that has a token, SPL token program or even some of the DEX infrastructure. And it makes cross program invocations a lot easier. I've been working with some teams for this hackathon, and wrote a bunch of cross program invocation examples for these teams to get up and running with PsyOptions pretty quickly. And it was just seamless for them to use our data structures and serialize it, de-serialize, because as long as we're all using the same framework, it works.

Anatoly (06:56):

Yeah. This was is my decision, so you can blame me, but I really didn't want to build a shitty framework. And until people started building on Solana, it was really hard to know, what do they need? I think it would've been worse if we built a bunch of code that nobody could build with, because it would've been incomplete. I'm going to say, it takes a lot of discipline to do that, versus laziness.

Tommy (07:22):

It makes sense to offload it on to the actual DAP developers. It's a different beast when you're programming the underlying system versus the actual just Solana runtime program. So it makes a lot of sense how Anchor came out and who really is leading it.

Anatoly (07:40):

Can you guys tell me what worked really well? Or what features or anything for any other Devs that when they're coming into building on Solana, what stuff actually feels like a superpower?

Taylor (07:52):

Well, one thing that's improved a ton is the SPL token program and how you manage the token accounts and whatnot. That's definitely something that a lot of the new developers on Solana don't have to deal with. But back then we were building into our UI the ability to have multiple accounts for the same SPL token and it was super frustrating and whatnot. So using those associated token accounts and other kind of, I guess you could call them rapper programs or things like that, that just improve the UX significantly. Understanding those and why they're there is pretty important when someone's getting started.

Tommy (08:30):

I think taking it a step further too, how does the associated token program work? And what's really under the hood is the program derived address. I put together some documentation for people starting to onboard to PsyOptions or related protocols. And I'm like, you need to read up all these Twitter threads, these documentations on PDAs, because there's just so many things you can do with a PDA that's very unique. You can get a mapping just to accounts, you can create a unique constraint. So for PsyOptions, there should never be... Right now, there's no reason to have more than one of the same option and the fungibility of those options are based on the expiration date, the strike price, the asset pair. And so we just have a PDA that is seated with those parameters and it creates this unique constraint.

Anatoly (09:24):

Oh, that's cool. So you encode the constraints as, basically hash it into the address.

Tommy (09:30):

Exactly.

Taylor (09:30):

Exactly.

Anatoly (09:31):

And the taker then has to satisfy those constraints to be able to take that trade.

Tommy (09:36):

No, not on the trade level, just on the general structure for creating the option. It's like, okay, if you want to spin up a BTC 70,000, USDC strike for the October 29th expiration, just that structure that creates that... because that's structure is the core structure of PsyOptions, the Psy American program. And that's what then controls the option TokenMint and writer TokenMint and how you dull out those option tokens. And so it's just there can never be more than one of those specific to those constraints. So it's separated from the trading concerns.

Anatoly (10:14):

Got it.

Taylor (10:15):

Yeah. I think you thinking of your stateless escrow. I thought that was a pretty cool proof of concept.

Anatoly (10:21):

Yeah. I wasn't sure that you guys already built... I think this idea has been around in crypto for a while, so I wouldn't be surprised if you guys use it too. But I like that idea that, because you don't want to generate infinite number of these markets, if everybody enters the same data, then it's going to spit out the same BTC month increment whatever, like May 2021 option or whatever you want.

Tommy (10:47):

Yeah. And we've seen it too. It's really useful to have these deterministic ways to look up an account address. So it's like, "Look, I can just check if this option market already exists by using these parameters, the PsyOptions market exists." And we also ran into some issues that we had to hack together, on the client side, because Serum doesn't have these kinds of constraints. And an adversary could come in and spin up multiple Serum markets for the same asset pair. And then when you're pulling that data from the chain on the client, it's like, "Well, which one is your UI using? Which one are these automated traders using? All that kind of stuff. How do you sync them up?"

And so that was a pain point, and we had to whip together a package. But now with Serum's permission markets and some other stuff, we can now use PDAs to say, "This is deterministically how the UI is going to determine the market. Here's how everyone else should do it. These are the seeds." And then it keeps everyone in line in a more decentralized way, rather than having to have some NPM package with metadata and it's painful to maintain.

Anatoly (11:54):

Got it. That's cool. What actually runs the market? Is it a Serum Q, a Serum V2 or V3 Q?

Tommy (12:03):

Serum V3 right now, for the Americans. Yeah.

Anatoly (12:05):

Awesome. Man, that's super cool. How was that integration? Is that blood, sweat and tears still, or are the tooling itself around Serum getting better?

Taylor (12:15):

It's getting better.

Tommy (12:16):

Blood, sweat and tears.

Taylor (12:17):

Yeah, but it was definitely blood, sweat and tears. I think that's what took us the longest part in the original hackathon that we won, was doing the Serum integration. And we weren't even doing any cross program invocations to Serum at that point, it was literally just client site integration. And that was really difficult. No documentation, got to read through the source code. I think we even found some bugs in their type script package and had to patch it ourselves. So yeah, definitely blood, sweat, and tears there.

Tommy (12:48):

There's still room for improvement. I'm like drop in list every time as hackathon participants start asking, or users are complaining about settling funds. I'm just at a constant stream of, "Hey, we should document this and add a flow chart for that." Because all the customer surface is offloaded to the people using the Serum stuff, so we get that inflow of feedback from users and other developers building on top.

Anatoly (13:12):

Yeah. People don't realize how strapped every team is.

Tommy (13:16):

I agree.

Anatoly (13:17):

It's literally like three, four engineers at best to, no customer service, no nothing, just pure software, open source software. It's not like when you look at a market cap of something, you think there's a equivalent to market cap S&P 500 company with 30,000 engineers just all cranking away. Thank God it's not, honestly.

Tommy (13:41):

Yeah.

Taylor (13:42):

Yeah. It's got its ups and downs. At least you can move fast, it's not a bureaucratic process. But at the same time, customer support definitely dwindles and I think 70% of people are probably testing in production. So the end users are just going to have to deal with that and understand that's just the way things are done in crypto right now.

Anatoly (14:03):

I guess, how close are you guys to launch and what are the next blockers?

Tommy (14:07):

So we actually are on Mainnet trading with BTC and ETH markets right now. We have been live since the end of August, just with BTC and ETH for the September strike. Then we upgraded to a V2 of our American protocol with Serum permission markets, so we can eventually close those markets. And so that gives us the ability to open a bunch more. And so we're live with those, we're working with a couple other partners to get some SOL markets up pretty soon. So we'll probably announce that here.

Anatoly (14:40):

Awesome. What have you guys seen in terms of adoption, and how are people using it and has anyone surprised you with what they're doing?

Tommy (14:49):

It's tough right now from the retail side using our user interface. I think what the biggest thing that I'm excited... There's been a lot of great feedback. Options are not an easy instrument to use, managing your own positions is tough. And so we've gotten a lot of great feedback from the community and it's shaping what some of these projects that are work thing on during this hackathon. I think that's what's most interesting and surprising is these teams that are building on top and they're not user interfaces. These are protocols that are going to be managing certain strategies and rolling positions for users, and so you can have this more passive product. It's like a ribbon finance to the basic ones, where it's just selling covered calls and secured puts or things that.

But there's a lot of plans, I don't want to leak their Alpha. But a lot of plans for additional products where it's more just, set it and forget it. And it has certain properties detailed out to hedge for various things, give you certain direction on volatility. And it'll make these... all these products, some more user friendly for retail, but also big institutions that are looking to hedge existing exposure.

Anatoly (16:00):

That sounds like you guys are building more of info level for options.

Tommy (16:04):

Yep.

Anatoly (16:05):

That's awesome.

Taylor (16:06):

Yeah. We chalk up the V1 American that we built as just a primitive, and as decentralize as possible. It doesn't rely on Oracles, it doesn't need pricing information. So the only dependency is the Solana runtime and SPL token program, I guess now Serum with the permission markets. But the original one had only SPL token as dependency.

Tommy (16:29):

Yeah. So there's capital inefficiency with the American style, because you can exercise at any time up until the expiration. So we're about to hopefully announce pretty soon, we have a European that we've architected and we're going to break ground on that and we'll crank it out pretty quickly. That will have a little bit more dependencies, but it'll be more capital efficient because it'll be auto exercised and we'll have a margining system built into it. And the American will continue on because we're going to build, I like to call it Carta for DeFi, but just a place where people... We whipped out a vesting contract the other week. And we'll be able to show people their tokens that are vesting, their options that are vesting, the ones that have currently vested and the options all in their portfolio and whether they should exercise them or not. It'd be less like trading based and more of just an interface for managing your portfolio of vesting stuff and options, so.

Anatoly (17:33):

That's awesome. How many engineers do you guys have?

Tommy (17:38):

We actually just hired another front-end guy today. So we're two full-time front-ends, and we hired another protocol developer, so we're two full-time protocol developers. Then we have a community guy and a marketing guy, and then couple of part-time and open source contributors.

Anatoly (17:52):

That's so small, I mean that's awesome. I feel this is the biggest thing in crypto, is how fast small teams can ship really sophisticated products.

Tommy (18:04):

Yeah. I think, as I've learned, the hardest thing nowadays or right now is, it's not the programming, it's the architecting the system to fit the runtime and developing the instruction set. And once you wrap your head around how that whole system works and you have your instruction set, writing the actual code is not that hard. If you actually take the time to just think and focus, and you have to have the knowledge and experience to understand that, it's pretty easy to start architecting a bunch of stuff and delegating and managing a little bit more.

Taylor (18:35):

The thing I will say on that though is that, the runtime changes here and there, but the changes aren't that drastic. But when you're using dependencies like Serum, Pyth, whatnot, those change a ton. And so you're seeing a ton of changes on Serum, so one week you might have architected something for Serum B3, sounds great. All of a sudden Serum updates to some new thing and that might change the optimal architecture for it. So you have to be nimble in order to just go with the flow as different protocols update, and as new versions come out and new architectures are viable.

Anatoly (19:14):

It's weird to think of immutable code still having dependencies. But something with Serum, you're so dependent on liquidity in those markets that if they move to V4, you have to update because you can't point to a empty market.

Tommy (19:29):

Yeah. We bring a lot the liquidity ourselves. Well, these are brand new markets that we spin up. It's not as much of a pain point, it's more just announcing and coordinating. But it's more of the European protocol and architecture, it depends on a lot of the stuff like... it doesn't depend heavily on the SPL token, contracts aren't represented as SPLs. And so it depends on this new architecture that they just announced, that Bonfida has been working on. So it's just interesting, you have to keep up to speed with what exists in the ecosystem, so you can constantly be like, "Is there an improvement? Can we squeeze something out of this is?"

Anatoly (20:03):

Is the European option, are you also planning for it to be Oracle free, or no Oracle?

Tommy (20:08):

No. We'll rely on an Oracle just for the exercise. We're wrapping up the architecture and probably just, we're going to develop this one totally open source from the scratch. I just put up the boiler plate repository and its open source. We're going to open source, or at least make public the architecture, so everyone can read and comment on it while we're just cranking it out in the next week and a half. So there's a Oracle dependency just on one instruction, just to actually lock in the index price, that would be for the expiration. But we don't see it being too risky of a dependency, considering it's not an instruction that has a lot going on so we can do a lot of checks. We could pull two different Oracles and reduce the potential pitfalls there.

Anatoly (20:53):

Yeah. This is a hard problem too. When an option is exercise it's still going to hit the Serum market to actually exercise the price?

Tommy (21:03):

No. So on the base layer, the European, it's just going to... essentially the architecture is locking in the price, and then users basically have to settle up the positions and collateral themselves.

Anatoly (21:14):

Got it.

Tommy (21:14):

The best way to describe this one is Deribit on chain. It's really just like P&L, not the full underlying.

Anatoly (21:22):

Okay. So you can actually settle in any collateral. You could have an option on SOL, but settle in wrapped ETH or whatever?

Tommy (21:29):

Well this one, it's actually going to be... well, it's going to settle in the currency that it's trading. So BTC, it's going to have this siloed market and account that holds all the BTC and manages the entire portfolio, margining for someone's BTC options. And so it has it's own realm of just, this is the BTC world. And everything settles in BTC, everything's traded in BTC and premiums are even in BTC, but then it just uses the USD index price to actually settle up on the strike. And then SOL would have its own world, with its own portfolio margins system. So they're not cross margined between all those at the moment.

Anatoly (22:10):

Got it. Is cross margining something you guys are also thinking about?

Tommy (22:15):

Yeah. It's one of those things where we want to just crank this out and ship fast, because it's improved from the existing architecture, for when it comes to a trading perspective. And then we'll discuss a more improved cross margining system.

Anatoly (22:27):

Do you think that there's a gap still in this idea that I think, what's popular on DeFi Ethereum is liquidity mining, and I just want to put my tokens and get yield? And is there a gap between that and options trading and central limit order books?

Taylor (22:46):

I think there's a knowledge gap. The closer you are to dealing with the primitives, the more knowledge you need to have, the more hands on you have to be in managing your positions and whatnot. So I think that reduces the addressable market or the end users that are willing to participate. And so that's why you have people building programs and tooling on up to manage the position, so it can be more passive. Because I think that's one of the biggest things that drove a ton of people to DeFi, is the passive yield, all the token incentive programs and whatnot. So I do think that there's a bit of a gap, but it's slowly being closed. And the more passive it can be, the more non crypto people or even crypto native people, but the less financially sophisticated you could say will come in and utilize DeFi.

Anatoly (23:39):

So you guys imagine that... or there's probably somebody already building this, where I have my token, I'm an LP, which is under the actual thing behind that position is a covered call or some other fancy strategy, iron condors or whatever, right?

Tommy (23:58):

Yeah. So there's a couple teams from the hackathon building that right now, actually.

Anatoly (24:03):

That's awesome.

Tommy (24:04):

That's what I'm really excited about. Because that's what we've seen is, there's decent order flow, I haven't looked at the volumes because we're just very focused on product. We know what the low hanging fruit is, so we're not focusing on the vanity metrics at the moment and not really talking about the TVL and whatnot. But it'll just increase order flow because these people can just get passive yield from covered call products, or they can hedge certain positions just by depositing tokens. And it's all going to be managing these underlying options and straddles and things like that.

Anatoly (24:38):

How long does it take to go from, let's say I wanted to build an iron condor or something that as a strategy, can I do that? Do you guys have examples already, reference implementations for things like that?

Tommy (24:53):

Are you talking as a protocol or as just a user, using the... like a client?

Anatoly (24:58):

As a, here's my DAP, I'll take tokens from LPs and then automatically generate the position on PsyOptions.

Tommy (25:07):

The hard part actually isn't to the generating the initial positions, the hard part is handling how they want to roll, if you're trying to do it over time, where they just can keep that open. So the generating the positions is super easy, placing the orders. We have examples, CPI examples in the repository for minting options, exercising, placing an order, opening a Serums open orders account, all that kind of stuff. Just been cranking out examples as people ask for them. And then, it's onto those teams to handle that really tough part of, how should we roll? There's certain concerns in there for manipulation. There's certain concerns for front running, there's certain concerns for eating through the order book and having to build your own TWAP into it and stuff like that, so.

Anatoly (25:59):

Yeah. Man, you guys are taking on some really tough challenges, that's cool. This is something that I wanted to get good at, trading. Trading options and deep learning into these things, but I got it to work.

Taylor (26:15):

Yeah. It's a full time job. That's why we try to focus on the primitive and lower layers and try to get that right. So then other teams can focus, if they're much more financially savvy or have of better trading backgrounds, can handle that. It's a full-time job to be a trader, to come up with those models, to build those positions and roll them, it takes a long time. And you constantly have to be updating them too.

Anatoly (26:44):

How long did you guys trade options before?

Taylor (26:47):

Not much. We're just retail traders. I interned at an investment bank once a while back, but to the extent of my full-time finance career, that was about it. And then we would trade options here and there, but nothing serious. And then, when we wanted to automate that option trading strategy, that would've been probably the first automated system we would've built. I don't think we built an automated option trading strategy before that.

Tommy (27:15):

Yeah. I would say we relate best with the retail, speculative, YOLO option users, rather than very sophisticated options traders. But it's been nice building this and winning that hackathon and getting some attention, because then those people show up. And we have some really smart TradFi people who have been around crypto, some really smart TradFi people who have never been around crypto, contributing to the thought leadership of where we should go, what's needed to get to certain structured products and things like that. And that's been super helpful because we've been early in Solana and have the engineering capabilities and knowledge to work with them of a translating their vision into a Solana architecture. And so we've just been helping as many teams as possible that have that background and can bring that knowledge. And then, that's why we're just like, "Look, we'll help you as much as we can because you're going to help us answer some of these questions that we don't know." So it's been good to fill out the team and the surrounding circles with that.

Anatoly (28:23):

Do you think that DeFi is something that... I always think of it as growing faster than TradFi versus replacing it. Do you think these products are good enough to compete with traditional finance, or are we just going to see more stuff being built on open finance because it's easier? I don't have to go talk to a CME to launch an option for my in-game bullets for my shooter game or something like that.

Tommy (28:52):

I think it will be just the fact that it's open and anyone can do it. Looking at the architecture here and designing an ideal architecture for the most capital efficiency system, it's just not really... You could do it in CeFi so much easier than you can do it in DeFi. I don't even know if it's truly possible. We're still just on the back burner trying to figure out how you could portfolio margin everything. I think a lot of teams that we've talked to are all thinking about that in the back burners. It's like, how do we margin against everything? So I think it's definitely moving faster. I think they will rival CeFi, a lot of these products, but I think they're still going to be both working hand-in-hand.

Taylor (29:42):

Yeah. I think both have their ups and downs. The speed that DeFi innovates because of the open source nature and things can be represented and it's all digitally native, it just makes the pace of innovation faster, also makes what you can build much faster. Like CeFi you're beholden to, not that you're not beholden to regulation in DeFi, but CeFi there's a lot more red tape. You got to jump through hoops in order to be able to launch a market or... You can't just launch your own equities exchange, it's takes tons of money and resources and whatnot. So it stifles innovation in that respect. So I think even if DeFi can't become as capital efficient as CeFi, you're still going to have more innovative products, more flexibility in what you can do with your assets, that at the end of the day, you might not need that capital efficient, high, super fast, low latency systems to do what you want to do with your assets. So I think there's a place for both. And I think DeFi is just going to continue to innovate and outpace growth in terms of TradFi.

Anatoly (31:00):

Well, our goal is to get that latency to be as low as physics allow, and then we're competitive.

Tommy (31:10):

That's why we're here.

Taylor (31:10):

Let's do it, man.

Anatoly (31:11):

Won't rest until we're building neutrino emitter detector. I just think with gaming especially, the first massive multiplayer games instantly within six months had a market for the digital items there. As soon as you get something like Star Atlas or equivalent, like World of Warcraft that's decentralized with all these assets on chain, I think the idea of options as a service, people are just going to, "Well, I got whatever... I got more gold that people want to use because this game is hot right now." People are going to definitely spin up those markets, it's just going to happen.

Tommy (31:48):

100%. We've been talking about game... we're gamers ourselves, and I haven't played a game since I really dove into Solana development 11, 12 months ago. But I'm hoping to get back to it once Star Atlas and Aurory and all those other... Kaiju cards, everyone starts actually launching the game play, I'll jump back to gaming. But we've been thinking about it a lot and what could be done with this American primitive, and that got us into talking to other teams and other games, just to see what's out there. And then I actually got connected with Metaplex and built out a contract that they just announced that's focused on gaming. And it all stemmed from trying to think about, these games, everyone's so early and not really thinking about how these game assets are going to plug and play into DeFi protocols and things like that. And there's still just so much work and research that needs to be done, and some infrastructure needs to be built for it all to work perfectly together.

Taylor (32:42):

Yeah. I think the interoperability for gaming is still... there's still going to be some rough edges there, because it's harder to build standards across games. But I think you'll have a few games come out and maybe they'll have transferability between games and whatnot, but it's going to take some time and some trial and error before we get to this on chain metaverse where you can transfer assets between different game worlds and whatnot. But I do think that is going to be one of the ultimate killer applications on blockchain.

Anatoly (33:19):

What are you guys excited about out of this hackathon?

Tommy (33:22):

Oh, for me, it's really just the stuff we've mentioned with the structured products, passive yield products, all that kind of stuff, being built on top of PysOptions. I'm very heads down on product and everything at the moment. So aside from the people that are ping me, asking me for help, I don't really know what else is being built.

Anatoly (33:42):

Yeah, likewise. I see NFTs being launched and then I'm deep in the trenches and optimizations. I guess that's good. It means that there's more stuff to do than you have time, so you started actually going heads down and working.

Tommy (34:00):

But there's a lot. The roadmap with just these teams alone, is ridiculous. We have so many products that we want to whip out on top, and hoping to launch the first few in the next week or two. And then the framework's there, it'll be a little bit easier.

Taylor (34:14):

Well, it's just fun to see people building on software that you've built. I'm sure you and the rest of the Solana team get excited as new projects come out and new people innovate. And I think that's one of the more fun things to do, is just sit and watch what people come up with because you can't come up with every idea yourself, so might as well open source stuff and have community run with it.

Anatoly (34:33):

For sure. What should we be looking out for? Do you guys have any announcements you want to leak?

Tommy (34:40):

Oh man. SOL options coming soon. Passive yield products that will make it extremely easy for people to get volatility exposure or generate yield. And this is yield that's not going to go away. A lot of these pools are based on rewards, and API is based on rewards and things that will dry up and aren't sustainable. But the volatility is a little bit more sustainable in a sense. Sure volatility will decrease over time, but.

Anatoly (35:09):

So these are like covered call strategies, basically?

Tommy (35:12):

The first few are the most simple covered call and secured put strategies. But then there's going to be a few other vaults coming out once these are launched, then it'll be focus on a few other vaults that have different strategies. Eventually, we started talking to some other teams like Symmetry, because we want to get a good crypto index, because then if we get liquid index options, we can create a nice volatility index for certain baskets. So that's all on the horizon. And so, you need that rolling and rebalancing and infrastructure, and that's what we've been working on the past few weeks, or other teams have been working on. And then it's formulas for just managing array of positions.

Anatoly (35:57):

What is your development process like? How do you guys go to build test and ship?

Taylor (36:03):

In terms of roadmap and how we determine what to work on, it's pretty ad hoc, things change up weekly, biweekly. But we try to run in two week sprints, at least just pick like, "Okay, what... base on user feedback, check GitHub for issues." Obviously anything that's blocking usage is number one priority. And then it's, "All right. What do we want to see built? What do our partner teams need, and how can we get them going?" I don't know Tommy, you have anything to add to that?

Tommy (36:35):

Yeah. Protocol development too. I'd sit and focus and start drafting up a full architecture doc with the instruction sets and potential functions that are needed, black box some stuff, just to make it a little bit easier and then put a to-do to dive in later. And then you have this whole instruction set and a general outline and framework, and you know it fits into the Solana runtime because you've made sure that the constraints are handled. And then I'll dive into running a test driven development process with Anchor, just doing full integration test. You end up writing a lot more test code, but I just find that the confidence level is so much higher. You can refactor and upgrade versions and you're just so of confident in your code when you have all those tests, so. And then, it cuts down the time from DevNet testing, where everyone just puts up a contract and then just relies on interacting it to test it. Especially when you're building a primitive, you want to have all those cases handled.

Anatoly (37:36):

In your use case specifically with options, what bug are the most worrisome? Is it overflow or actual logic and economic?

Tommy (37:45):

Probably logic and economics. The American protocol, overflow is not an issue, not really. We do all the check map of course, but it's not an issue. Maybe if we get some weird [Altcoins 00:37:58] eventually trading, then we'll have some weird issues. But I'd really just say logic, economic attacks, things like that, when we get into the capital efficient Europeans that has the margining system built into the base layer, and liquidation built in the base layer. And just always thinking about account management. You have that limitation of number of accounts you can pass in, and how to architect around that.

Anatoly (38:19):

So if Solana could change one thing, what would it be? Or anything, and things-

Taylor (38:24):

Two things.

Anatoly (38:29):

... two things. Finite number

Taylor (38:29):

Fixed account length. If we could make it-

Tommy (38:32):

Oh, sure yeah.

Taylor (38:33):

... dynamic sized, I think that would be great.

Anatoly (38:36):

The length of the data.

Taylor (38:37):

Yeah.

Anatoly (38:37):

Okay.

Taylor (38:38):

The account data.

Anatoly (38:39):

That's actually... I don't know if you guys saw, but believe-re growing reallocation from the program itself or the account that it owns, I think it might be live already in 1.8.

Tommy (38:49):

Yeah. I saw ...

Taylor (38:51):

I think I saw the PR for that.

Anatoly (38:52):

Okay.

Taylor (38:53):

But that's one thing that... Sorry, but when people jump over from ETH to Solana, that's probably the biggest gotcha, that we're like, "Oh crap. I can't readjust or create a larger array, more mapping data, whatever." So that's one thing. And then also the number of accounts you can pass to an instruction [crosstalk 00:39:14].

Anatoly (39:13):

In a transaction. Yeah.

Taylor (39:15):

... open up more. Yeah.

Tommy (39:17):

And I think 1.8 handles a lot of these headaches, but you still, when you're trying to think, for the long term, just the limitations in general. I'm assuming they're always going to be there, the number of accounts you can pass in, can't be-

Anatoly (39:32):

109, the goal is to double the transaction size basically. So the number of bytes that a transaction can maximum size. That means you can double the user data or encode more, put more accounts in there. So there's always a limit because of the real time nature of the system. You're not submitting an arbitrary large transaction that then the block producer decides, "Okay, I'm going to pick this one." You're really like, "How do I write to the block right now?" And making sure that doesn't slow everything down, is a challenge. But really cool, man, you guys are shipping like crazy. It's awesome. It blows my mind that you guys were hackathon team that is now... there's teams in the hackathon building on top of PsyOptions.

Tommy (40:23):

I love it.

Anatoly (40:24):

Yeah.

Tommy (40:24):

I've been doing office hours, every Tuesday and Thursday, just letting people come in and ask questions because it's just nice to see people building on top. And we're going to do whatever we can to help them out and keep them there.

Anatoly (40:36):

That's super cool. Man, really good to catch up with you. Thank you for coming on the podcast. Is there anything you want to add for the listeners in the final bit?

Tommy (40:46):

Yeah. I would say, check out PsyOptions to trade your BTC and ETH right now, SOL coming soon. And then we'll be announcing an under collateralized European protocol pretty shortly, going to try and crank that out as quickly as possible.

Taylor (41:00):

Yeah. And get in touch. There's no shortage of projects that we can dream of and I'm sure others are too, but happy to help any team out that we can.

Tommy (41:08):

Yeah. And if you're a protocol too, looking to do option liquidity mining with American PsyOptions, reward contributors with options, or use the PsyOptions vesting contract, we're trying to get that. The vesting contract's a unique one, where you can delay your vest. The recipient has the option to delay their vest if the issuer grants it. So that way they can keep pumping the vesting and the potential taxable event. Not an accountant, so don't take that tax advice. Not financial advice.

Anatoly (41:39):

Not accounting advice, not financial advice. That's awesome.

Tommy (41:42):

Yeah.

Taylor (41:42):

No advice.

Anatoly (41:44):

That's super cool. Well, thank you guys.

Tommy (41:46):

Thank you.

Taylor (41:47):

Thanks for having us

Hosted by Simplecast, an AdsWizz company. See pcm.adswizz.com for information about our collection and use of personal data for advertising.