Mert Mumtaz on Pudgy Penguins, Solana's Vision, and The Blockchain Endgame

Transcript

where where did that name come from man and uh we learned about the launch like date roughly one day before and then I messaged my co-founders go [ __ ] and when did you learn that pengu was dropping on salana and and what was your first reaction welcome back to the rollup your home for technical talks and open conversations that shape the modular future nothing on today's show is to be construed as Financial advice you can see our full disclosures list at tholl up.co ventures sit back relax and enjoy today's episode after a quick word from some of our awesome Partners imagine if the world's GDP moved on chain guys say GM to safet the newest protocol by safe isn't a blockchain and it's not a new layer 2 it's a transaction processing Network

built to send money anywhere and swap anything on thousands of layer twos from a single Universal balance in sub 500 milliseconds it's fast it's smooth and it's simple and it's built with safe multi-sig level security in mind safet unlock smart accounts across multiple chains allowing users to supercharge security scalability and speed so get started today at net. safe. Global and remember to follow at safe onx what's going on guys welcome back to the rollup today we have mertus Prime on the Pod murus I don't know mimus Prime what the hell where where did that name come from

man I don't quite remember people like to make derivatives of my name mert because it's just a very it's one cable and some guy called me murus prime or murus Decimus Meridius or some something Roman or Transformers so I just go with it nice yeah man thanks for thanks for joining it's your first time on um we've been we've been infamously a modular podcast but we are we've came to learn that the whole world is modular and that the idea of modular is really just kind of a marketing meme um pinned up against the monolithic as you could say salana so we'll talk about that but we're definitely not going to make this into another Mao versus monolithic

debate I I think we have a lot of agreement actually on ideas um about kind of how to scale the ecosystem and uh kind of where where we align but before we get into the tech stuff man like you're Canadian my parents are Canadian I'm Canadian I also dropped out of college uh I went to the University of Florida for two years started an online business just dropped out you dropped out as well so I mean it that was a really weird time back in my life I mean I was leaving University something completely different um you know feeling anxious feeling like I don't know if I should do this like just that whole path and I mean I'm curious about your story with that like you know what was your experience going through

University and then dropping out what did your friends support you like how was that experience for you and H how has it shaped you because I know it shaped me yeah M mine is slightly different because I finished undergrad I dropped out of grad school so I went to a school called Queens University in Canada and it's a pretty small one but also had some solid crypto Talent so my friend Lucas Manuel Charles St Louie for example are some of the earliest people in maker and my co-founders for helias Nick and Liam they also went TW in fact Liam was my roommate for five years and

Nick was also in the same program we did and that program was it's a it's called Apple math but it's applied math and a double in Computing Communications engineering and Communications engineering is basically signals analysis radios stuff like that control theory and I I went into that because I didn't know what to do uh in Queen's first year it's a general engineering program program and you don't have to pick a discipline I actually wanted to be a civil engineer at first and then I was like you know what I'm going to be a theoretical physicist and so I just did math so I thought that would help at the most but also the engineering side would help you with instrumentation and things

like that and that was going well I was always aiming for like perfect gpas and stuff and getting them and then I did an internship at Blackberry on the cloud infrastructure team and this was a cloud responsible for supporting BBM where all those messages back in the day used to happen and that was a lot of fun and much more fun than school and so when I finished that internship which was roughly 13 months I came back and I was like you know like [ __ ] school I'm just going to I'm just going to ship stuff and focus on business and and commerce and you know whatever I can just read books to learn about physics if I really want and so I did that and then

graduated worked at a bunch of big banks in Canada and startups and then I had like this I I really want to get into drones because I was like you know the future of warfare I don't know why I was thinking this but dude that is such a pressing topic right now as well yeah I was like I was like as a kid I used to always do this which is like okay what's going to be really popular in the military such that if I mastered it now I would be useful in the next 10 20 years and roughly four years years ago I was like okay that that's probably drones in robotics so I got accepted into Georgia Tech masters in robotics and I did that for a little bit and then I was like this is so

depressing like this is not how you learn I just want to ship [ __ ] and whatnot and then I was still working on my day job which is a startup that does fintech and one day there was a um I was responsible for building a payment system to go from Canada to Australia and it was super inefficient it it had to jump through like different jurisdictions and basically I said why not just use crypto here right because I was already familiar with crypto the whole time and they said it was a scam and I did not like the answer and so what I did was like okay I'm just going to join coinbase which was hiring at the time and I was roughly their first or

second Canadian hire and while I was at coinbase I was like okay if I really want to master crypto I can't do both crypto and Robotics so I just dropped out of the the Masters program did your did did did your parents or anybody close tell you to stay were they were they at all disappointed because on my end I had my mom was like no no no no no you got to stay I mean I I was undergrad right it was different story I was I was 18 when I dropped out and then my dad was like nah dude [ __ ] it like just floor it you know like no sh like just if you know if it doesn't work out just just go back yeah I didn't I didn't really so my mom passed when I was young so I have my dad

but the thing with my dad is that he started he dropped out of the the one before high school I don't know what it's called in English just Elementary School uh grade school he dropped out of that at like eight years old and he just started hairdressing like he was he's been a hairdresser since 8 years old and so he definitely whenever he tries talking about school I'm kind of like that shut up like you you didn't even go to high school and so he didn't care he he just had trust um so it wasn't as bad for me yeah yeah yeah yeah it's different it's different when you get through undergrad um well cool man um yeah I think I I just always I always

appreciate people who who are able to kind of take that leap I think like I think there there's something different between those who who leave school behind uh in terms of Entrepreneurship I'm not sure if it's like a back against the wall thing or like a you know you've kind of erased part of your plan B um but it definitely kind of inspires people to to act in a certain way and I mean we we all know some of the famous entrepreneurs in the world are you know dropouts of jobs universities High School Etc so um yeah man let's get to the meet and potatoes of this conversation starting off with penu um I'm a Peggy penguin h i a disclosure I've been invested since like two years ago in the seed and been actively helping the team and all that

good stuff and your pfp is a pengu now um which is a really sick pengu actually um so what is your first reaction when you learned that and when did you learn that pengu was dropping on salana and and what was your first reaction I learned about it probably a few months ago and at the time I believe Luca was still evaluating his options and it was middle of the night for me like 2: a.m. and he's like M and he had talked to somebody else from Foundation Pedro and he was like so we're thinking about doing this like do you have some time uh either next week or now and I was like sure [ __ ] it and so we jumped on a call and at first they

were debating like whether to launch on salana or not and like if you do it on salana what do you need to do to ensure success like who do you talk to um all these things so I helped them with you know some introductions ideas on who to air drop to obviously the infrastructure with helias and all these things and I was pretty impressed with how fast it came together and I was talking to Luca about it and he's like wow like this is how you guys do stuff on salana which to reference how speedy everything was and be because like we shifted plans a few times it it was going to use ZK compression at first and then it was

like okay wait a minute like but you need to the ownership of a penguin on ethereum and so a claim might make more sense do we do this in two phases oh wait Jupiter doesn't have support for ZK okay so then we were like wait should we get support for ZK from Jupiter or talk to Phantom and so there was a lot of missing or uh moving pieces and uh it came together pretty fast and then yeah it launched what was it two days ago now um and it was very chaotic at first there was a lot of things that went wrong on multiple fronts that we can talk about but overall turned out well yeah I think that um some of the people in the community were curious as to why they

didn't wait for abstract um but I think now that that mayet is what like at least a month out or so um they want to ticket this out and then Al I mean the the the beautiful knock on effect is that this these communities kind of come together right like I don't think you'd be rocking a ethereum nft bro like that's just I don't think you would ever do that but then now it's like you know what the penguin just makes sense and so like these small things right they I think they make a difference in like unifying people broadly um which which I'm excited about I actually haven't even claimed uh I'm a Believer I I'm not a Jeter on day one two um but I like the reason I didn't claim was because like I just everyone was struggling so like

what happened there like I saw your tweets you were going on a rampage about it um uh yeah I mean I think some of the majer guys are framing this like oh it's just another example of Sal's congestion problems and like this thing is not going to be able to scale like what happened there from a technical perspective and is this something that salana Community needs to be concerned about or it could have have been optimized better what happened yeah so well first actually to tackle your point on like some people were wondering why this wasn't La launched on abstract I think that would have been very dangerous that you just launched this L2 it hasn't been battle tested yet now you're giving it a huge event like you could have just broken it

or got the fees super expensive before the team really had time to iterate and it might have given off a bad first impression of the chain so I think it actually is smarter to do it on a different chain first because you can just bridge to it in terms of what happened with the claim yeah so I know basically everything that happened here in rough sequence so first is that um the Clusters API which is what you needed to um it's it's just a web 2 API that actually I believe got dosed or it basically just broke so it it simply wasn't working for a while um metamask marked the site as suspicious or as fishing and then uh so

those were like two pretty big ones and then obviously they were getting dosed and then there was also some things that weren't communicated properly on the front end so like um for larger Penguins or like larger allocations the transactions were split into a few transactions and so then people would be like wait a minute um this doesn't show my full allocation well it's because not all the transactions happened yet and so there were some UI things like that and the the the salon specific issue um is roughly and I I tweet about this is it's it's basically just like what happened is they're obviously not salon devs right they're they're notm devs

they're coming over from ethereum and I I think they bait roughly started the salon stuff like two days before the launch which I was like guys please let's let's let's pick it up here like we'll happy we're happy to do dry testing and all this stuff and uh we learned about the launch like date roughly one day before and then I messaged my co-founder like oh [ __ ] uh get get ready for some trouble um and like we were asking for their code like we're like okay we'll just write the code for you and stuff and we never got to do that but roughly what happened is um two twofold one doesn't have to do with the blockchain which is on salana to you know add that screen where you have to connect your wallet and then

pick the right wallets it's just a simple SDK it's the web3 wallet connector and I believe they didn't set that up correctly so only Phantom worked but then like Leisure didn't work Soul flare didn't work backpack didn't work and if you try to connect with those you would just say off error and people thought this was like a salana thing but like when you just connect your wallet the blockchain hasn't been initiated in any way yet that's just a pure wallet interaction and that's just from the SDK so that and I think that's still broken if I'm if I'm right unless you use Phantom uh and then okay the second one that's uh more chain related is block uh uh transaction expired errors okay so this is an interesting one so the way it

works on salana is transactions have an expiry date um because you fetch the latest block cash and then you use that for your transaction and then within like a minute it'll expire so if it doesn't get accepted within that one minute it's not going to work like it'll just expire and you'll have to reconstruct a transaction and the the way also it works on salana is there's paralyzation and there's also the concept of compute unit limits per the block um which is kind of similar to gas on ethereum but obviously different so the the the way works is there's a block and you can fit 48 million compute units into it okay each transaction consumes a certain amount of compute units right so

a very very simple transfer might be very few compute units a very complex one like a Jupiter aggregation routing through multiple dexes might be much more complex um and so but then the other thing and this is the important part here to understand is that since salana is split into multiple different accounts right so maybe there's a penguin account there's a Jupiter account all these things there is a smaller limit per account right so there's a 12 minute uh 12 million compute unit limit that you can write to On Any Given account and so you take that there's a 12 million cap and then you take the fact that when they were constructing transactions they were not

setting the compute units properly because they had one day to write the code so they would request the max possible compute units for the trans transactions so in this case it was 200,000 Compu units and then there was 12 million available because they didn't like Shard the account structure properly uh because they just didn't know right it's not it's not like and perhaps that's like the knock on salana that I think would be pretty fair actually which is that developing on it is just harder right like this isn't like very obvious stuff you need to really read the docs and work with somebody to understand this but like they just didn't have time and so what ends up happening is you have 200,000 compute units trying to fitting into 12 million compute units and that equates

to like 60 claims per second and since you have let's say a few thousand people trying to claim all at the same time obviously some of those don't make it uh into the blockchain before they expire and so we saw this as they started the launch and we were like oh [ __ ] these guys are oversetting the compute units by like 6X and so we gave him the code sample to fix set and then they were also not like setting fees properly um and then the third thing was they were not retrying the transactions so on salana if you send a transaction it doesn't get accepted you would retry it but they didn't have that knowledge and so once we add of those three things um which

was hard because they were also fixing like a thousand other things that weren't to do with salana um then you saw their Landing rate increased to 100% And so I think like a good criticism of salana there would be the developer experience is complex and this isn't something that just works out of the box from just using the S and you need to actually work with um people ideally for like you know a week or something before launching um and so that's what happened yeah I mean okay so lots of lots of kind of unpacked there uh but just to kind of finish off the lore then so you guys were able to then provide them with this code sample um with the updated fixes and then they just able to

uh plug that into like the repo and then press live and then from that point forward it was it was it was good to go y got it okay um and that was about like maybe an hour or two after the lunch so we I mean we gave him the sample a while ago but they were just getting bombarded by like hey metamask is fishing or claiming your compromised and like a ,000 other problems that they had and so I was like spamming Luca like dude like you needs to prioritize this one so that your transactions can land and after like I I don't I didn't look at the

clock but probably like two hours or something it was added and then once it was added it was fine yeah and so okay so to your original point and then we'll kind of get into the the the architecture of the blocks a bit because that's a particularly interesting point there um it this could have worked for a different use case with some sort of like ZK ZK compression Network extension concept if there was a better if Jupiter supported that uh that like so basically what so basically at the current state Jupiter does not support Jupiter uh deck uh aggregator cannot uh support network extension uh operations or like basically why can why could this not

technically why could this not have worked out like why can why can't Jupiter support um ZK compression for this claim I understand that we had to link to an ethereum wallet so that was part of it but like strictly on the on the salana side is that something that that Jupiter just needs to implement or like how does that specifically work um I mean Jupiter can easily implement it it's just not like priority uh yeah the the biggest thing was obviously you have to prove ownership of a penguin on ethereum and then you get uh so like it would have been two different kind of code paths that you had to builds that are different and so that would have been confusing for the users um but then they also wanted to cover the cost for

the users um and they also didn't have time right like doing this like okay so if they had let's say a month to plan for this this could have been all worked out super efficiently and all this but they had like a few days and so that that's actually the main problem um and then the rest of it is kind of just like Downstream effects of not having time right like if you first of all you wouldn't have structured a smart contract in that way um but even if you did you would have optimized a transaction sending unit which also didn't happen um but then let's say you did both of those then you still have like some rough systems designed math there right like assume you're G to get 5,000 claims per second assume salana does 3,000 transactions per second at

Peak um then the 2,000 people that are excluded from that what's going to happen to them and so what you would do is something like a q system where you would put them in like a Q and then actually add durable nonsense to the transactions or you would do a regular airdrop right you would take a snapshot you would have these addresses and then you would just use so salana has like a Merkel tree distributor um or Z compression to just knock those transactions out um and that's what like Jupe did that's what jto did and that's what all salana teams do and that's why those go well but the result of like an evm team coming over for the first time and trying to learn it right away uh I think it was just too chaotic of a combo

yeah yeah so so then uh would is there a clear use case then for like these Network extensions or ZK compression for these highly targeted events with when there's going to be a lot of onchain transactions or are you still of the mind like no let's just all put these on the Bas layer the base layer can handle it we just got to build it right well so Network extensions is certainly an Overlord term um and I I hate that term personally I'm trying not to call them rollups to not piss you off for for ZK compression um it's the execution the transactions are still down on the L1 um the only thing that's different is that the state

availability is um uh hashed together into the latest kind of account hash that you linked and so to replay data for example you would replay the the the The Ledger right so like something like call data on ethereum um and with with ZK the the the thing that you that ZK compression helps with on salana which is so this concept doesn't exist on ethereum which is why it's like hard to make sense of which is that salana has additional rent costs for storing data so let's say I want to send you a token right now uh and you just came over to salana and you don't have any token accounts right you just have sold the main coin

but let's say I want to send you usdc well if I want to send you usdc but you don't have a usdc account I need to pay rent so that I can allocate a usdc account for you now that rent is uh refundable meaning that once that transaction is done and that account is empty I can close that account and then get the money back but like that's an additional cost and it doesn't matter for one transfer but if you're doing 10 million transactions then you're not only paying for 10 million transaction costs you're also paying for 10 million new account costs and so what ZK compression does is it condenses all those accounts into one account so that you're only paying the account cost once and so you're able to save a ton of money but the execution is still done on

the L1 like the transaction still happen on the L1 got it cool um I mean all in all it post that little hiccup the the quite a successful launch I mean tokens trading quite High obviously in terms of fdv market cap um also significant amount of float which I thought is particularly exciting because I mean it's just a change in what we've seen um and then I imagine for abstract what they're going to do is just use like a wormhole or some sort of like native token transfer and just basically put a liquidity pool there and the two homes for people will be on uh salana and and abstract and maybe even ethereum L1 but like for all intense purposes man each

time that I've onboarded somebody this holiday season I've just been sending out everybody like a hundred bucks in fart coin and they think it's like the greatest thing since slice sprad and and it's like yep download Phantom sign in with your email and here's a 100 bucks of fart coin it's like the easiest onboarding ever so I think the thesis for um pengu coin on Salon as far as like the uh proliferation is working I think there's already 500,000 holders um so yeah I'm curious if you have any comments on like the overall tokenomics there maybe that plays into a broader shift of like these you know Venture back projects or just kind of any closing thoughts on what you expect for penu for the token for the community on saana um and just kind of like broader

broader takeaways outside of kind of the slight Tech uh hiccups I I didn't really look at the tokenomics too much I know Luca wanted to make it super fair and actually get a lot of people involved and not do anything predatory like that was a big Focus for him and so that's obviously and and he also obviously he's a product thinker like he doesn't care okay you know about religious things he cares about like okay where are normies where can they go today and one of those things and uh you know top two things let's say is salana along with perhaps base um and so it's like okay um so that that seems like it was a thesis and

then what what it means I like I have in claims mine uh I I do like the Penguins quite a bit and so I don't plan on selling unless like it just goes really up and I'm like okay you know I'll take some profit um but I think it's a good unifying force I actually really like the abstract team for example um actually like roughly everybody on their team that I know uh I already follow them right like Jared or Saar or um whoever else so I like the way they think and I think we need more of that thinking and so that's probably also another reason for launching on salana because it's a relatively similar

philosophy in how people think about users and and and experiences um which is a good segue into into another shift uh which is we we did a pod with Max Resnick in Bangkok talking a lot about the ethereum road map dilemas things things of that nature and I think a lot of the stems down to some of the points that you're making here with regards to being product focused versus you know like ideologically focused or aligned um and I think so Salon has done a good job of of building this in increase bandwidth reduce latency product I ideology um and so I just want to kind of like touch on your your first you know I think it's maybe been a week or something but I'm

sure that they've been working with Max for uh several weeks or months now uh what is Max coming over to salana building uh Anza mean for the community and like what is Anza trying to accomplish that's going to help you guys with salana and with helus yeah so you can roughly think of Anza as consensus but on salana um instead of ethereum and Max actually I had him on light speed this morning and what he is working on I believe they called them like lead Economist or something so that's a fun title and there's a few things that he'll work on and those range roughly from defining consensus properly to

better fee markets to me and then perhaps like Dynamic issuance or if issuance changing actually even makes sense um so things along those lines in terms of what Anza does it's um they maintain the uh the valer client that's the salon Labs clients or it's called Agave now um and it's it's an interesting dynamic because soof far answer is coming out soon and that'll be the second client and it's a contrast to ethereum where you have I believe four clients or maybe even five but I think four evenly split um versus on salana it's more like you have two and then those two kind of compete with each other almost where so one of the things that people one of the narratives

that push is like well once fire dancer comes along it'll be a million TPS and everybody will switch to it and so what's used for Agave but that's actually not correct uh because the Agave team is extremely talented and they will compete to uh keep up with fire D and they might even do a better job honestly um so like they're already raising block limits they're already doing they already just set up a research team as well so you're GNA have kind of this nice little competitive Dynamic of like both clients trying to improve uh against the other one but then you also have the nicer piece of Mind where the likelihood of a bug occurring on both their clients at the same time because they're in different languages and all this uh all

the these things is is low so like you have better decentralization and security and redundancy or just peace of mind um and yeah so that's that's roughly what and then Anza does other things like they also work on you know um specific svm stuff like svm apis research um ZK stuff like that as well yeah yeah I mean I think that uh that I that like concept of just uh the competition between the two clients and then also just fighting uh fighting it out to deliver the best performance is pretty aligned with with Max's idea for ethereum as well uh I think I think the

Justin Drake presentation might have sent him over the edge uh but I think it was he he was already bubbling up based on our podcast in Bangkok um I also just think that that culturally speaking um you know Max and others there seems to be a a strongly different culture in salana then in ethereum in Celestia in Bitcoin um like product focused um very very application oriented um and so I think that that actually made a made it for a good fit for for Max per se because ethereum kind of has is in this is in between a rock and a hard place kind of right where there's a lot of infrastructure that's been funded to and we keep the there's just been this

creation of new problems that is then solved by more infrastructure that is basically creating more problems um but but I think that like the two there's clearly a difference in where the like the scalability comes from in these two ecosystems and I want to talk about that actually with Celestia and even Ripple to some extent um but just let's nail down this in increased bandwidth reduce latency concept um it we have Sonic has just launched on Manet 10K TPS I believe on evm is what they're claiming monads pushing the evm both in these parallelized highly functional databases using like w and customized DBS um Salon is all in on this increased band lth reduced latency idea um so

let's talk about what that means in in the immediate term and then looking forward with the additional uh client a fire dancer and Beyond um I think it's very clear that that people are in agreement that okay if you want to onboard people you bring them to salana it's very fast it's cheap um it Works uh I believe that if you want to get uh salana to a maximum scale where you want where you want to take this thing you know obviously we we need to scale it further there's trade-offs that happen here um and so I'm curious like what is the right mental model for okay understanding sana's vision for where it is today and now and this purpose that it serves and then what is the vision for down the line five years 10 years

and Beyond um where we need to to accept tradeoffs or at least uh uh acknowledge them and build towards a more uh censorship resistant and kind of maximally decentralized ecosystem while maintaining this core Vision so what is the what is the here and now why is this so important for you guys and then what is the kind of five to 10 year trajectory do you believe looks like to kind of keep this Vision intact while also navigating some of the tradeoff space well everything roughly starts with totally idea of NASDAQ decentralized NASDAQ now it's a silly saying because it's like te if

you take it literally it's a ridiculous concept but it is a good heuristic or a mental model for what you want it to somewhat look like which is to say that if you have markets on chain and if you believe that crypto or blockchains enable global permissionless markets and real-time price Discovery and um you know the facilitation of value and and how it moves around the world um then it just turns out that if you optimize for syncing information as fast as possible around the globe on kind of one machine or or one state machine not one literal machine um then as a byproduct you also get other things like lower latency and increased throughput which tends to be

better ux for people but everything comes from that core concept of just syncing information as fast as possible around the globe including sensorship resistance for example right um and realtime sensorship resistance status right like if you spin up a node and you send transactions and the blockchain's job is to duplicate that as fast as possible around the world then that gives you a lot of sens of resistance to begin with um now of course it's not perfect there's a lot of um the the way that I've heard this um and I used to say this um before I discover that actually the former CTO of coin uh salana said it who's Greg is like I've always thought about it as more of like

an engineering versus science approach where like so both can work right so you can do science and you can uh you know conduct experiments you can do some peer reviews and you can kind of work your way upwards from there that that approach works or you can do the engineering way which is like you ship something you get feedback once you get feedback you iterate and then you have uh new kind of directions open to you so one approaches uh one example of this is like when salana launched it didn't have fee markets right uh not quite sure why but it just didn't I think they like really rushed the launch um like most of the parameters were even just straight

up set from Cosmos like the inflation schedule is just straight up copy pasted from Cosmos um because some of the earlier valuators were Cosmos validators and but you know uh it turns out that if you do it the engineering way there's some tradeoffs such as like short-term let's say uh uh reliability issues which was very common in 2021 and 2022 but actually since um threee markets in quick went live it's been almost two years now without an actually performance related outage there was one outage in February of this year and that was like the silliest outage ever

because it was already something caught on testnet and everybody already knew existed and it was just like a matter of devops of getting it deployed but like if you look at the performance characteristics none nothing has that like that has actually um affected the change since now I think it will in the future but it hasn't um in that twoyear time span and so in terms of what it looks like today so very simply put it's not really about increasing transactions per second like TPS is just an easy to understand metric and it can be used as like a good Northstar if you say something like 1 million um but the actual idea is always composable Global

Capital markets right that doesn't sound just succinct as decentralized NASDAQ but that's what it is and those markets can be you know shitcoins but they can also be rwas they can be stable coins they can be uh Soul they can be deepend coins they can be anything right it doesn't matter we don't the network is not opinionated about that um they can be jpegs and as you kind of work through that Vision you start noticing a lot of issues in production so one of those might be me for example right and then okay so what's the long-term fix for that well it's probably something like

multiple concurrent proposers right where you have a bunch of different leaders around the world and then you have competition amongst the block leaders um and obviously as a user you now have multiple options as a leader and you can pick the one that's not malicious but that also has the other effect of um reducing latency right because now your uh transaction doesn't have to go from Singapore to New York you can go from Singapore to Singapore right because there's maybe a leader there and you know we can get into the blockchain trilemma and whatnot and why I think it's generally a super outdated model um but fundamentally there's a lot of engineering problems that are just not being looked at like Bottoms Up

optimizations and it's not just about you know one of the biggest misconceptions is while salana just tuned up the hot Hardware requirements and it's like well you know EOS did that Finance smart chain did that poly everybody has done that and they did not get close to Salon scale and that's because that's not enough right in in fact if I'm to put it succinctly sala's actual design principle is informed by toi's years of working on phones and phone operating systems right so how how do phones work you let's say optimize this chip uh and the software that goes with it and by the time you release it there's a new phone that came out now you have to rewrite it again and then

there's a new phone that comes out and you have to rewrite it again there's a new ship set there's always that and it's a similar concept to blockchains which is to say that the the goal of salana is not to jack up Hardware requirements it's to write software in such a way that it gets out of the way of Hardware right the end state of Salon what that looks like is and like when people say you just add hardware and the change scales that's not how it works the second you add Hardware there's like 5,000 different new bottlenecks that arise that's why you can't just do that right it's it's even comparable on ethereum you can't just you know boost the [ __ ] out of the sequencer because okay now you're Bing on da right so there's always different components and

so the question becomes how can you get to such a state where the only thing you need to do is if there is market demand for this block space let's say it's 2 million TPS um Can the node operators actually just go on Amazon buy a new Ram stick or something and then actually meet that demand and the economics should make sense obviously right the cost that what you make from handle that extra demand actually offsets what it costs you to set up that node so that's it's a little confusing but write software so that it doesn't get in the way of Hardware is the actual design principle yeah okay so um in the in the in the immediate term um we have we have fired answer uh

rolling out as as as a second client and and we have a clearly an an ecosystem that uh is is functioning at scale I mean uh there's no there's been no other ecosystem that has generated the amount of of transactions and user activity in terms of either meme coins or trading or whatever it is defi since I mean since the Inception of of all this kind of onchain work then Salon in the last you know year or so um that's not really what I'm concerned with like I you know I get it like we can get this thing to really high levels of scale for the current user base but we're also entering like two separate propositions

for salana at the same time first one is AI agents in a world of an agentic onchain reality these these agents are going to be able to push more transactions and at a larger scale than I think any of us can even imagine so I while I think that there might there might be an over supply of block space now I think we'll see chains that are fully dominated by agents and I I expect to see um a AI agent transactions happening on salana uh in mass in 2020 uh then the other element is like kind of comes back to like like maybe like something like penu like if penu gets to a point where where they're onboarding you know hundreds of thousands of people on on a daily basis or a weekly basis

monthly basis to salana to to trade um that combined with a bull market combined with a a gentic future I'm struggling to see how we scale salon's base layer in a in a unified censorship resistant way to to to create this decentralized NASDAQ long term without needing these things like ZK compression or these basically horizontal scaling so so while you talked in depth there about the like the now how do we handle this perceived future of of higher demand for crypto for Native assets on salana uh as well as just broad AI agents like is is

just is this fire dancer is the addition of this client going to work uh is that just going to be enough like like what do you for foresee um and are you just going to be like against horizontal scaling because we can just do all the execution on the Bas layer or is there something is is there a different future that you may see here I'm just trying to pick a bone because I think that like it's an important conversation Redstone is def's modular Oracle supporting over 100 clients on over 50 chains offering 1,000 assets Redstone is the leader for Bitcoin staking and yield bearing assets they offer the most robust customizable price feeds available in any model join the

Redstone exhibition [Music] today welcome aboard on our journey into the singularity this is the endgame road map for fra and fraxel the new L2 out of the fra ecosystem that is totally based on fractal scaling fractal is a modular L2 blockchain built by fracks with modular and fractal scaling in mind fractal scaling is based on the mandal BR set of fractal images and it explains how l2s become l3s and the entire app chain life cycle goes through its

Evolution the FRX ecosystem is made of some of the top builders in the space and has stemmed from its OG days of the fra stable coin it is now in the LST space the LRT space and this fraxel L2 joins the existing Suite of fra products including Frack Swap and fra lend you can see frax fairy the bridge to go from any chain arbitrum optimism or main net into the fra main net Bridge your tokens and bring your FRA assets into the fractal ecosystem join today to get the best out of fraxel you are still early

join today take part in our Expedition into the singularity so I would say like it there's actually not too much of a difference in in how let's say the rollup Centric road map scales versus sou salana scales it it's you can actually basically do the exact same thing on both the question is whether that's integrated within the same system or whether it's you know outsourced uh or or uh the parts are you know it's it's kind of like if you contrast building an Apple computer to a Windows computer right with apple you don't have choices you just you have to use their unified memory you have to use their

chips you have to use all these things with the windows it's like well maybe you can get some part from Dell some part from Intel some part from some other thing and maybe that's analogous to like use ethereum for the computer part maybe use uh an L2 for the ram part execution and then maybe use a DA layer for that's custom you know between Celestia Avil ethereum migan whatever um but you you don't fundamentally ever get around the actual limits uh and and the the ultimate limit is bandwidth like it doesn't matter what architexture you have that is always going to be the limit and so like for example uh let's take BAS okay so base is in L2 and it helps scale execution uh on ethereum

because you you take a bunch of transactions you do them on base you compress them back you roll them back up and then you post them on ethereum um now the the problem is okay what happens when Bas gets overloaded right because fundamentally now Bas needs to keep increasing gas limits okay um but base does not have for example local fee markets which is to say that let's say in your AI agents example let's say one of the agents is particularly popular right that particular agent is now going to make the fees for everybody else that are trying to transact on that chain higher and so okay what does Bas have to do well they have to I they have they can do two things and I asked this Jesse

on my podcast they can either isolate the hotpots directly on the sequencer right so local fee markets paralyzation or they can use l3s right uh which is because l3s are basically the same concept where you have you're offloading again you're kind of sharding in a sense it's not actual sharding um which is fine that's I think that's valid um but you can also do that without needing an L2 or an L3 right so if you have the local free markets and the extensions directly on the L1 that's also the same thing and so the difference becomes and I think this is actually the Crux of the modular versus monolithic thing which is I think Sovereign block space that's that's what

I really like about modularity which or which is to say like it's not necessarily monolithic versus modular it's actually much more about app Centric versus general purpose Centric right because you can have modular with different l1s right like the cosmos uh or common Weare from Patrick from Avalanche and the the the one thing is like okay what use case and and so you must work backwards from the use cases so imagine you are an indie developer a game developer and you want to build something that you know posts game transactions on chain at some interval you don't really need salana

for that right like what is salana it's a giant Global State machine it doesn't really have any of the properties that you really care about you can easily use your own app chain you can use your L3 you can use your L2 that's app specific you can use your app you can use your all one um and so then it becomes a um an exercise of when you work backwards from the use cases which solution actually best fits the business needs of that business right if it's something like order books for example perhaps it's better on salana or perhaps it's better on L2 um but then now you're G have the same problems on the L2 which is the execution hotspot problems um but then also you're GNA have the same sens

resistance problems and the way to get around that is to actually have a bunch of different nodes everywhere uh and maybe multiple conent proposers um such that you end up actually kind of arriving at the same problems right and so and this is actually captured in at a high level actually by balic blog endgame which is to say that perhaps most blockchains actually look pretty similar in the endgame um the question is how do you get there and that question is extremely difficult to answer because it's completely dependent on the use cases until whatever that game looks like so for example assume that um uh uh let's say nfts are like the next big thing for the next five

years just as an example perhaps they just all go to abstract chain right and perhaps that's like 90% of all crypto users today okay it doesn't matter if Salon is superior at that point in any scalable fashion because the users are using the nfts on abstract chain okay and so by the time that everybody gets to the end State then abstractors already built a mo around State and and in distribution right and so that's that's the actual question to me because the scaling part actually Nets out to being pretty similar on all chains and I think within like five 10 years all chains will be scalable anyways uh and they will have methods of dealing with elastic demand whether it's near L1

sharting whether it's ethereum's L2 uh road map or whether it's Salon as kind of state machine or S State machine approach so I think they all work the actual question is a business question in my view which is which use cases and what what do those use cases need yeah I think we we are in agreement that that there's a broad convergence here of uh design philosophy at the end of the day I mean it looks somewhere in between in my view somewhere in between like uh where where Celestia and Salon are like heading and now one thing that that you put out which I thought was triggering for the ethereum community maybe

is that you think that Celestia will have will basically have a large amount of growth of rollups or I think it was like win a lot of rollups next year or kind of grow their rollup ecosystem potentially now unlikely more than ethereum but that they they'll have a strong run um and I know that we we also are in agreement here that this purpose built is how I like to call these chains now or these purpose built or use Cas specific domain specific execution environments are and Sovereign block spaces is is kind of how you called it enables this customization that you don't get in the general purpose environments and and and I think that's one thing that is is super important for scaling um applications to Max

throughput as well as just for differentiation of use cases right some financialized use cases have different security preferences gaming Etc there's just a lot of different use cases that can be built on different Tech and I think I mean it's honestly like like the onus is on us for education and on Builders to be able to understand the technology and build it where they think it's the best fit but know coming back to this this specific tweet I I view this like modular thing as partially being like okay we build out specific uh use case chains so celesti in this case is strictly just for da and eventually for verifying ZK proofs um to enable interop um do like like like I guess the the question is

what triggered that tweet and how does that tweet with regards to Celestia's potential growth for their rollup ecosystem how does that relate to to uh ethereum and salana like like like what triggered that and how do you see these three kind of players in a in a basket of kind of use cases or uh what the what the vision is um for kind of what they're trying to build so what what triggered that tweet is I always put a business or Market lens on everything uh and then combine that with the tech as opposed to kind of just working out where from the tech and I was basically thinking okay so

what needs to happen on ethereum for let's say sufficient value capture is well first blobs need to enter price Discovery and then the all twos need to capture users um right so those two things need to happen but then those l2's like okay assume you want to build a rollup today let's say you have an app and you want to build it into a rollup okay so what do you do uh do you Deploy on base well perhaps but then you're gonna have the same problem as salana where you're kind of leaking some value right okay so then what do you do do you launch an all2 on ethereum and it's like well perhaps if you really care about the assets on ethereum right and you you want trust minimized uh uh bridging or something and extending those assets to the L2 but if you look at it from a

Market's perspective right there's a lot of al2s on ethereum that are very strong there's arbitrum there's optimism there's a bunch of [ __ ] on optimism there's ZK sync there's Stark nut there's all these different things and so then it's like okay wait a minute um and by the way all those guys have roughly different incentives like they have some unifying incentives such that like it's scale ethereum and all these things but from the business perspective they have different incentives which is why interoperability seems to be harder than it actually is because fundamentally capitalism kicks in and you've seen kind of uh vitalic tweet about this right and so then if you're if if you're an app and you want to build an app chain okay so what other options do you have well perhaps you can

build an L1 but that's a whole lot of [ __ ] and takes a lot of effort or you can go to the chains that are purpose built for this reason where it's only exists for this reason there's no other kind of weird things to distract your users or have kind of misaligned incentives where um uh let's say this other L2 wants a different interop standard and then polygon wants DK egg layer and all these different things and so on Celestia it's just like you you don't want to as a business Builder you don't want to go to an already big market right which is a counterintuitive thing you want to go to something that is relatively small but on the rise up in fast growing and so then if I'm to

build a rollup I would probably go on Celestia right because first of all I can work closely with the Celestia team because they need to support these guys um because they are behind um ethereum but then also they have a much more Clear Vision for helping rollups versus ethereum also does but it's also bundled with some other things and then like then there there's the argument like okay if the alto's main and this is like what John Shard would say which is like alus are good on ethereum because ethereum has a lot of valuable assets on the L1 and you want to be able to kind of extend these assets okay where else is that the case Bitcoin right in fact it's the most valuable asset and so perhaps then you think of extending

these assets to enable better Finance well you can also go to bitcoin and because what's the alternative the alternative is you compete on some level with arbitrum and base and that's hard right nobody just starting out probably wants to do that so that was kind of what triggered it for me and then like it really depends on the type of rollup right uh I actually have been a a huge fan of BAS roll ups since the day that the Forum post went live because I tweeted something about like uh sequencer censorship and then code replied to me and linked me the post and I was like oh what the [ __ ] uh and then it was it came out that day and so I was like wait but this this would also make

quite a lot of sense on a blockchain with very fast block times right like so Salon is currently 400 milliseconds but it's going go to 120 milliseconds and so you can have a really good experience where there's a Basse roll up on salana and the sequencer is the chain and it's super fast anyways and so then it becomes the same problem that I just described in the previous answer which is what are you building and what requirements do you need to satisfy right so yeah that's yeah so but but but isn't the problem for salana uh with with regards to like a base roll construct or like the reason why that that you guys have built the ZK compression because the da like Sal is just not meant for da basically um so would that actually work

or or like how like because we have bullet and Ellipsis are two of The prominent ones I think Zeta Market also has launched in this kind of structure differing structures but for my understanding the the bottleneck for salana like for rollups is or these types of like constructs is the da like what's the how does that how does how does that work uh well for bullet which is a The Zeta team um they're they're using Sovereign Labs um and actually they're using us so we're working with Sovereign Labs um to like okay so to see how much we can push this thing like when you are posting um these different blobs to the L1 what does that look like

how fast can you get it and salana is not designed for like da right uh at least in the way that something like Celestia is optimized for it but since it's optimized the other thing which is syncing information as fast possible it already has a pretty good da layer on its own even with like basically no focus on it right it's kind of similar to ethereum where before um blobs were added people would just store it in the call data and like that still worked now it wasn't the perfect solution but it still worked but since Salon is already much more stealable than ethereum the base there then the call data approach actually also works on salana um and the the difference here is going to be

the core devs right like on ethereum if they're like wait a minute our customers are rollups and so what we need to do is we need to build the L1 in such a way to serve as uh the main supplier for these rollups versus salana is like no like we our main customers are not rollups our main customers are people who want to transact and have execution here uh and so we're not going to build anything purpose-built on the L1 level just to support these rollups but if you think it's useful to uh use our fast block times for something like a Bas scoll up then perhaps that might make sense yep and I think that's like the big that is the big question mark for ethereum it seems like very clearly Bitcoin is very clearly this store of value this gold digital gold it's very clearly U you

know maybe they build out a ZK verifier on the L1 and a trust minimized bridge and that we get op cat and you know we get a different ecosystem in the future but but for now it's very clearly like what it is salana is also just very clear execution oriented let's scale this on a composable Bas layer Celestia is very clearly opinionated in the sense of like we are just building the Bas layer just for rollups the most optimized Bas layer possible single slot finality Etc and then ethereum's kind of like in this place where it's kind of trying to do a little bit of everything of all three of those and so I it's going to be particularly interesting and I mean there's been endless podcast about this conversation as of recent so we don't need to get into it too much but like it's going to be interesting

over the next couple years to see how how ethereum kind of fights this little war that it's in um of basically positioning and like what what engineering resources need to be allocated where because basically I I think ethereum will end up trying to do kind of all three but it won't ever be able to do all three in my view as well as those who just focus on the one I mean that's just a very clear uh idea like specialization wins one thing I'll point out and you know people think I'm like super anti- ethereum or something that's not the case I'm anti certain uh social layer industrial complex people from ethereum but the chain itself is just technology I have nothing else piece of technology um and so there's actually an argument to be made for fighting all

three battles because like in business what might happen is if you focus super like you can take Intel for an example in uh uh when they they so they Intel used to build memory chips um but then the Japanese people started outperforming the [ __ ] out of them on memory chips right but then they were already kind of somewhat doing microprocessors in the back end and then they're like wait a minute why don't we just use microprocessors and then really scale that out and so there's also something to be said about fighting multiple battles because we don't know what the next year two years three years n years looks like and so there might arise an opportunity where that chaotic approach might actually yield better

results because you have more stuff to work with versus like let's say the future is just like nobody wants to do execution and all1 for some reason now I don't think that that's likely but just give as as an example well in that case what's Salon going to do right that was their one bullet in the chamber and so now there needs to be a pivot and while I don't think like while I think the the the the verticals are set properly such that you know Bitcoin goes to Value Celestia goes for da and then Sal goes for execution it might be the case that there is some combination of those that ethereum will still do much better um so I I don't actually think it's all bad yeah I mean it's prob hedged and I mean they will they have such an e has such

an advantage in every one of those categories effectively in terms and then on top of that users and liquidity um but nonetheless it'll be it'll be interesting to see which which of these components the market values the most over the coming years and then how each of these teams specifically kind of lay out their their future groundwork for that reality um yeah and that's the beauty of being in in in a market um cool okay so let's pick on salana some more um and talk about Mev which you've already done uh at naam on Twitter and on uh light speed I I you you had a really good pod with toly which you were

kind of the the question asker but but it's becoming more and more apparent uh in different realities and I think one for me that's hit is like not setting the slippage perfectly on Jupiter or uh accidentally doing like Auto on like a low cap meme coin and then getting getting hit with like 10% sandwich bot if you look at Dex screener the order book is filled with sandwich Bots the reality is we're also calling this a mechanism of value for salana we're we're using this as a way to basically um describe the network effects of salon's L1 now one thing in particular that you've been bringing up

is this his idea of rev which frankly I'm still trying to wrap my head around this a bit so maybe you could just give us a a quick lure of like the meev on salana and then what what is this rev concept and uh how is helus involved in this and just yeah what's what's the latest here and um should should the community be worried about uh the situation of me on Sal yeah so you you can think of R which is real economic value let's say as basically how much money the validators make and as a result how much the stakers and the token holders make uh and the L1 right so basically revenue is is kind of how you think about it now the way it works

on salana is um as a validator you can make money in three ways one is obviously just issuance right comment all proof of statement chains number two is priority fees right so people there's uh they want to land their transactions prioritize them so they pay small fee for that or there's me tips right so jeto tips and the interesting part about the jeto tips is that like 80% of them are actually just for landing faster right or like sniping a liquidity pool uh I believe ghosts which is like an indexing team did an analys on this where it was roughly like 8 to 9% 7%

something in that that range of the me tips were for sandwiches and so if you take that into the whole context sandwich um the sandwich contribution to Total rev is single digits like it's it's it's relatively small now that's not to say it should exist but it is to just put in context um and so what's been happening um is these TG Bots right like the I'm not going to name them but there's a bunch what what happens is um you have a shitcoin okay and that shitcoin just launched and people want to uh uh buy that shitcoin as early as possible but the problem is the shitcoin

is like a 50k market cap okay and so that is to say the price of the shitcoin might go up by like 100% it might go down by 100% And it usually does um by the time that your transaction even makes the blockchain and as a result if your slippage is like super tight on a shitcoin it's it's the same thing as like a reverted transaction on evbm right like it makes it the chain but then you don't want it to execute because it would be at a ridiculous price okay so that's like almost all the sandwiching happens from those TG Bots and those users just setting a wide slippage so they can just buy the coin because the math that they're doing is like I don't give a [ __ ] if I get like the worst execution by 20% because I'm

buying the [ __ ] in the hopes that it goes up 5 billion x right that's the math they're doing um which you know like there's certainly a lot of improvements there to be made um now the recent me talks that have happened everybody is getting this wrong uh because nobody understands how M on salana works so what's happened is so ethereum has a a mempool by default on the L1 salana does not and so what used to be the case a year and a half ago is jto who does me infrastructure it's an it's a validator client that all validators run to boost their revenue it's like me boost on ethereum for example kind

of um they had a global mempool so they basically added a meol to salana and sandwiching started picking up then and so what jeto did and I was roughly me and John sh were like roughly the only ones against this um jeto shut down the meol and I was like oh [ __ ] why did you do that like that's just going to make sure that some bad actor gets a private pool going and then starts sandwiching retail but then none of the L one actually benefits from it like it's not democratized so that worked for a little bit like I don't know six months maybe five months and then what happened is uh there's a bunch of dark pools on salana where they have like these side deals with these validators um or like

these rich whales from Asia or something and they leak transactions and they sandwich people but then the revenue from that it doesn't go to rev so this is where everybody gets it wrong that sandwich revenue does not go to R because it's not in protocol this is completely a side deal so on in addition to the Rev there's these dark poool sandwiches that take those profits from those slippage uh from the sandwich tax and then distribute it to like this set of let's say five validators I don't know the exact number um which is bad because if that keeps happening then that pool keeps getting bigger and bigger until it's centralized effect and then it just compounds now the question

that I've been asking is like okay what's the rate of growth right if let's assume these people are 0.3% of the network today so helus is our one validator and we are 3.26% of the network in terms of stake and that's roughly 3.1 billion in so uh in in USD and so then the question is okay these dark pools how fast are they growing are they going to become a centralizing Force in months or five years and nobody has a data on that and actually the data seems to suggest that it's actually going to be a very long time such that that's a problem um does that make sense so far yes it does I'll let you continue but yeah we're all good here for for

this initial point just some questions about like whether these dark pools are are basically then Distributing the revenue directly to themselves effectively like their own no that that they're running yeah okay so that's why it doesn't contribute basically to the RV yeah so like the the interesting part of this is that like there's a very easy way to eliminate it which is jto enables the men pullback because since 90 you know x% plus of the network already runs jeto then most of the network would also it would democratize the sandwich just so everybody would extract the same amount and so actually total R would go up quite a bit yes but it would come at the expense

more sandwiching whoever's doing this yes well they would also lose the money these dark pools would also lose the profits and then there would be more sandwiching on the main on on any other transactions because they would just be doing it on the on on everybody basically or on every transaction that hit any validator right and so why does goo just turn this back on because of this reason then yeah so because then the math is something like okay um is the rate of growth or is a central ation concerned from this small dark pool worth everybody getting sandwiched more on the network right that's there's that's not an objective question yeah yeah and and so and to answer that properly you need to really

understand what's going on with the rate of growth um and you know that's still a discussion that's being had but then also there is a bunch of other mitigations that are possible on the app level right one of the things that um certain people on SAA would say and that would be me included is like we also don't want to like just straight up give up this fast because we know what happens when you play this out it ends up looking something like ethereum in its current state which is not which is not like horrible or bad by any means but it's also perhaps not ideal or if it's maybe there's a better way right and so then it looks starts to look like so Ellipsis Labs which is also launching in L2 release the sandwich resistant amm and so okay what if radium now uses that

and then pump uses radium and so they use this new thing and so maybe that reduces quite a bit okay what happens when there's a team called deflow and they have something called conditional liquidity which is kind of like intense based trading that would also reduce it at the opat quite a bit and then what happens if the transaction Pipeline on salana which still has quite a few bugs is optimized such that you actually can land the transaction that fast before the price moves so you can actually reduce the slippage at the app layer and so you do those three things but then you take into account you can just enable the M Pool back on um and then the final state of course of this is multiple concurrent proposers right where you uh incentivize competition at the Block leader level and because right now there's no competition the L1 uh

leader they might sandwich because they just they hav unb sand uh the bundles or something um but if there's competition and the block times are reduced that starts to also uh uh really minimize Meb or malicious Meb quite a bit and so it's certainly a very very Dynamic and chaotic problem but the one thing that I take took problem with let's say is people conflating this with reev r just refers to Priority fees plus jeto tips but most of those jeto tips are not toward sandwiching and the sandwiches the actual problem there is they're making money to the exclusion of everybody else such that they could become a centralization concern later yeah which is a a massive massive issue but the

question then becomes how big of an issue is it compared to the reduction in in ux as for everyday users across the entire chain as well as the other kind of slew of dynamics that that you laid out um well what would be particularly helpful here would be uh Clarity into the dashboard or into the information as to what that centralizing Force really is extrapolated over time with varying degrees of like growth or Decay basically um and and so being able to to kind of see some of that information I think would help the community make a make a better decision ultimately jeto and then would would this jeto uh would this be a vote a dow thing a governance

thing or would this be like a strict engineering thing it would I think it'll depend on some business considerations and legal considerations on their end so I don't know the answer to that I mean obviously you can do it either way but like what minimizes the externalities for goo team right that's the question um based on so we are running the numbers on this and we should have like a report pretty soon based on what we've seen it it seems like the rate of growth is pretty slow like it would take a very long time and keep in mind that everything about the market would have to remain constant and nobody would have to make any nobody would make changes for this problem to keep going and I don't see that happening so like um it's

not as bad as like it's me is a problem and it needs to be mitigated but it's not like this thing that will destroy the chain within five months for example it's more like okay if this keeps going unmitigated on all fronts app protocol education all this [ __ ] data then yeah it's a problem but that's not gonna happen right like we're on top of it yep good man good stuff uh really great conversation I've really enjoyed this and appreciate you coming out um I've I've been super super excited about salana for the last I mean really turned turned my head maybe like a year ago or so uh and gotten much more involved during 2024 and um been very pleased

with the user experience and just the community growth and what's been going on excited to kind of have this have this vision of pushing execution to the max um and really trying to see what what we can do with that and I think one big thing there is the is the beauty of crypto and Defi and web 3 is like this composability so maintaining that at scale um I I think that's that's a very important part of the salon div vision and something that ethereum Celestia and these other kind of more modular or more like roll up horizontally scaled ecosystems are are working to figure out as well but ultimately I tend to agree on the convergence of of design um really really good points on pangu I

appreciate the the uh ex explanation there as to what's been going on I think the the community would appreciate that um and yeah man looking forward to running this back appreciate you coming out awesome yeah thanks for having me on Cheers buddy