Interview with Scroll - Pragma Denver 2024

[Music] so got two more talks to go the next up is an interview with Sandy from scroll and we're going to talk about what it means and how the whole process of launching in new2 has been like so please join me and welcoming Sandy on stage thanks for having me not a not a new um not a new face not not a new face to the east Global audience but this is now more of a a public catchup um that's also being recorded well the first question I have is um I found a typo on the scroll's website no how is that possible with how do I open a pull request on your GitHub correct this um I find that surprising after thousands of people have been trying to help us correct tyos and well we'll get to contributions on on open source maybe at the end of this but uh let's kind of start off by uh maybe giving people an overview of uh how scroll came to be yeah and you just launched main that just a few months ago it obviously kind of not even realizing zms would be possible two years ago to having one of many live and ready and actively being used um it sounds like a wild right to me so um maybe a quick intro on sort of how you started scroll and um and what was the motivation back then yeah of course um we started about three years ago and um at that point I've been in the crypto Community for a while as a contributor Community Builder and also sometimes investor and um I got deeper deeper into the ethereum infr layer research rabbit hole and um and I met my co-founders on the eth Research Forum um back then um y was a PhD student running a paper on uh DK proar and hen was doing AI um in AWS and um he also did a ton of research in compilers and um rware systems and so we kind of quickly came together with the thesis that you could build a CK evm that is by Cold level equivalent and um bring to developers something that is um very good developer experience but also very future proof so something that we can continue to iterate and um kind of move the ethereum BR map forward by scaling with a a kind of an a balanced view of security versus scaling and so that's we we started off as a fun kind of side project and quick snow FL from there that's awesome um given how fast the space moves not only just idea generation but also the execution of those ideas what does it mean to be future prooof yeah that's a great question I think at multiple layers of the stack from a technical standpoint um there are Innovations coming out of every corner and particularly now I think more and more researchers and kind of great Engineers are coming into the space so that execution speed is accelerating and the quality of CK teams in the space I can see a market improvement from what I saw three years ago and so that's that's very exciting and it also brings forth a lot of opportunities to collaborate I think um and on the ecosystem front things move even faster um I think the the the you know what people tend to ignore is the that the price sentiment and Community sentiment is a great accelerator for technical Innovation so the part of the GitHub contributor kind of meme that kind of took place organically in our ecosystem um for for those that don't know um a lot of our GitHub contributors are um kind of open- source contributors and they received a celesia airdrop and then later on the stockboard air drop and so it's been identified as a hot geub to contri contribute to and I'm sure it will become the destination for many other airdrops in the future um but that kind of um er raised a a question of like what is valuable contribution how should contribution be accounted for and um and and and that's kind of a a really fun space as we can try to shape where contribution goes forward so there's been a lot of um obviously like typo type of requests but also a ton of um researchers and also Engineers coming forth and saying do you have any open research problems where we can help with and um and and that's always the start of a great conversation where we could scope out some open problems where the core team doesn't really have any bandwidth to solve but they they would essentially become that that kind of community-driven build process makes the protocol future proof um yeah that the sheer amount of eyes that are looking to make sure that potential attack points or just improvements are actually on track um that's a really interesting and like a good way of looking at it um we've had a lot of l2s come up in the last few uh few years um they've definitely grown we're finally got a St where we can actually say look my my app has anywhere from 10 to 100,000 Maus which was a sentence unheard of three years ago and even less before that and the kind of the set of questions I have are about just maybe let's have a more honest discussion around what does it mean to launch a a lay 2 uh independent of what tech Powers it and uh and you're in that very early phases of kind of seeing that from the test net to building to test net to main net uh to where it is now and uh kind of the first kind of question I have is uh do your developers care if it's a ZK evm or if there's some new hot thing on on this VM compared to the other one um that's a great question I think um I think this ties into the fact of how we communicate the product to developers so on one hand there are the type of Builders and developers who are um very sensitive to um incentives and there are type of developers who are very sensitive to Vibes and there are the type of developers who are very sensitive to um cool technology that are being developed and um looking into the future seeing what kind of tools and design spaces opening up so that they create something new and so I think um with the recent phenomenon of this like multi- instances of L2 um I think that ugates and obviously kind of draws attention for a lot of people who um who wants to experiment with different token mechanism designs and I think that's there's a lot of interesting observations and lessons there so we're we're keenly looking at what's happening in this space and on the other hand I think it just kind of brings truth to the view that what is an L2 and what is the role of an L2 team so from our perspective the L2 team's role within the entire blockchain Community is to continue making bold technical bets and find directions where we think makes sense where the developers need to go and um that place may or may not be related to like more block space and more like non-interoperable block space um it might be interoperable or non-interoperable uh so not interoperable so um no one wants to have an experience of so I already have multi-sig kind of Wallets on multiple l2s and um and and that requires an Excel to manage my um personal kind of shitcoin portfolio and that is not a preferable user experience and I can't foresee that happening for hundreds of l2s in the future and um and I think so Innovation on the user end and Innovation on the interoperability end is required to push forward the next kind of wave of innovation I think um some like kind of more monolithic l1s can make bold technical bets and um it's very clear who's responsible it whose responsibility it is to kind of push the envelope and and find the next generation of build space design space but in the theum space the you know there's the good thing about being very decentralized and then the bad thing is you know I think it's unclear like whose responsibility it is to kind of keep pushing the boundaries of evm and keep building things that expands the developer designed universe so I think we're we're we're always kind of thinking about where does our Builders need to get to the next kind of million users and and then get to the next kind of million TBL and how can we be the most supportive in terms of giving them the best chances of succeeding I mean I completely agree like that that that's like the the obvious downside of what we see in fragmentation like this is perfect example of what fragmentation is and now you kind of see that at all levels of not just user experience but also protocol design um for scroll what does that actually translate to in terms of what you get to anyone from talk about to spend your engineering time on on on improving this so that you actually get to attract more developers is there an obvious answer you found since mainnet or over the last just year over over feedback or is this something that you think is just long-term commoditized so it may not actually be the differentiator and you have to think about other ways to attract users and and Builders um so I think I've had two recent observations as we're kind of onboarding developers and and helping to build a community the first thing is um there is very very strong appetite for truly Innovative um projects whether it's in defi or gamei and um when when when when users are looking across all the l2s there's a severe lack of like things to do so I think of l2s as a bunch of like thin parks that are now open but the rides are all the same and so whenever there's a new ride um opening um it kind of Demands like significant and overwhelming interest and and then that's great and and that should provide the kind of catalyst and incent for application Level Builders to truly innovate and um the second observation is that um I think um most users are still like very very price sensitive so uh we we we keep kind of um we we just dro the bridge gas fee by 50% and we're continuously working um to to to make the chain cheaper and faster more secure and more decentralized so these things we we have kind of um lots and lots of updates in the coming months um both kind of related to 4844 and and then also our internal kind of optimizations to kind of make the chain cheaper but just the cost is a constant complaint within the community and I can understand where everyone's coming from and um I think you know the the chain that can really make a leap in solving that problem while providing a very high degree of security Integrity is going to be very very very future proof so we want to be that place we want to be the kind of brand where we we maintain like you know high level security and um and and and gradually descend on the on on on the cost Factor so that's a very very top of mind for us I think on in some point we always knew that you know people are price sensitive but I think um what is just kind of having constant like real time feedback gives us it's also once you have a main net you can actually see their sensitivity differ across regions because 10 cents here may not be enough or much but it definitely moves the needle for somebody in some other part of the world and and you get to see that basically very real time yeah yeah yeah correct on the topic of also future proof from a from a tech standpoint um how do you think about talking to users and understanding their needs and using that to effectively shape and Define the protocol versus having a more neutral opinion on what will just support all these things and everybody else has the ability to customize it versus like hey like you need this thing like guess this feature is missing let us kind of figure out how we can work on this and introduce that as like a big protocol upgrade um have you gone into those types of conversations yet or is that an active thought um I think in terms of user feedback there are always some very uh obvious and uh very clear feedback um and there are some feedback that is um that we need to take the um take the initiative and um and kind of push in a SE Direction with or without kind of constant prompting um I think the the obvious ones are um people would like to be incentivized for onchain Behavior that's a very consistent and obvious feedback and um another feedback is that people are very Pride sensitive so they want things to be cheaper and faster that's also very well taken uh very very well like you know um taken taken note of and uh and and then the the ones that users currently are not super sensitive about it's about kind of what about the next user what about the next user so most of our onchain activities coming from relatively smaller wallets and they're they're they're relatively active so I think it looks organic in terms of like whale distribution um I think scroll is relatively like percentage of whales on scroll is relatively low compar to perhaps other chain and um and I think that's a that that that is sign of a vibrant it's a good behavior yes I think that yeah I think that's a good behavior and um and I think um um forgot where I was going with this um shaping up the the tax tag because of the request yes yes of course so um so so we need to think about if if that is the the the the kind of the community profile then how do we on board the next set of community set who are um who are perhaps not on crypto yet and um and that opens a whole lot of like Last Mile design problems where we need to collaborate with wallet and we need to kind of um think about this like interoperability problem so there's a there's a lot of um Innovation whether like you know you think of it may not be on the protocol level but um like once you kind of go to users there's a ton of kind of product level decisions that you need to make and you need to collaborate with multiple parties along the um along the user experience chain um which I imagine is a lot longer than your average and that's actually a constant cycle too because you can't be done on that one um I mean I was going to just generally ask uh which you sort of par already answer is like who whose responsibility is to kind of think about the abstractions because uh sometimes we've seen the protocol say well we're neutral we're not going to have you anybody decide but some people take a very active role and uh it seems like you're on the more active side on making sure that you get to solve the user uh user issues here um how do you think about interoperability altogether uh what does that mean to you and and uh what is scr doing to help simplify that um so I think there are interoperability is a very very challenging problem there are lots of technical hurdles to provide like a truly interoperable experience right now you can't do nft swaps across l2s and you can't there's a there's just a lot of things you can't do even within the e ecosystems even though everyone is evm and um and and that is a really difficult thing to do so I think the way we approach problems like this is think about kind of in in in in 5 to 10 years time if the if the ecosystem is to survive what would the end user experience look like it would be an experience where you could on board with a wallet and not worry too much about the chain underneath but somehow kind of have a be able to build a position and then experience and so when we start from that place in mind and try to kind of work backwards and in that kind of user journey I think there's a lot of explanation about why like AA wallet hasn't really taken taken off yet and um and also why kind of the the the the multi-layer kind of ecosystem positioning is actually really hard for coordination efforts one thing I want to come back to is so you kind of talked about the user Behavior you kind of talk the developer Behavior but but as a protocol how do you think about the geographic focus of where your community is and should be and and have you noticed any Trends or patterns or things that you think are worth optimizing for um so I think um I think we might be in the minority here um who thinks kind of um like Regional um Regional relationships and and and and dominance is very defensible and um it is of our view that if you are are the the the kind of the most if you build a community in certain regions and those are very longlasting and um and and and those kind of solid communities within kind of the more um unexpected places um Can can really help you build a brand and um and that can create kind of a crossover impacts into other parts of the ecosystem so um our our our experience building the community in um in Asia and also in in in turkey has brought a lot of valuable lessons and um and I think just the the kind of the Builder awareness and where that intersects with certain problems within like a particular subset I think those are uh those are very very um um like valuable and um but obviously I think you know crypto exists predominantly on Twitter and um Twitter is an now forecaster and now forecaster and um maybe later on something else and so um you know I think the the crypto core always exists somewhere online or on chain and um and um but um having kind of different kind of regional um grasp of Builder preferences and um kind of um building mind share um I think that's a that's that's a that's that's a strategy that we would we would um we think it's it's is fantastic and is playing out for us I think that that is definitely one of the the opinions that I feel like have uh fall in the minority category for sure but it seems like you've already seen an abration here with this and uh these things compound very fast so I can't wait for a few more months of uh data points on this my last kind of question to you is what can we look forward to from the scroll team over the next few months or this year anything interesting that's in the works or things you're excited about um I think we're just very excited to help some of our ecosystem project with their cold stars and launch um a bunch of them have been preparing for a long time and I think now is the um the optimal time for some of these kind of well well prepared projects to kind of go to market and um and in terms of our our core team um we constantly iterating and shipping and um and I think this year you can expect a lot of updates that can probably bring our overall cost and speed like up another order of magnitude so our Coe team is very focused on the long-term mission bringing the most secure and also the cheapest al2 to to our users and um and then we're also going to be thinking about the centralizing different parts of the stack starting with um puber decentralization and later on perhaps moving to um like sequence or decentralization probably next year amazing that's a that's a great set of Milestones to have and uh thank you so much for for this chat thank you for having me excited to have you again