EthCC 2: Rick Dudley

we thought we Hey good morning everyone me too wait okay I have the I just wanted this is awesome I really like these slides so far can you also do a third line with a @af Dudley zero a duuude ele y yeah y0 awesome yeah that works for me behind next to the all right hi everyone my name is Rick Dudley as it says up there it's really early here which means it's 3:30 a.m. where I live so I'm a little I'm always a little out of it when I'm here in Paris the title of the talk is pretty much what the talk is about so I guess there's two main themes to this talk and my previous talk a ttcc one of them is that there extemporaneous like I don't I don't plan so you know we're here in this famous space has all this historical significance where things are done in a very European style for those of you watching and some other time in some other place there's I'm in a basement I'm in a deep hole there's all these chairs so I'm supposed to be like prepared and like all these peoples was this watch but I'm actually much more interested in having an interactive or dynamic discussion so I'm gonna have everybody sort of ask their questions I have some questions from Twitter I'm gonna answer those and then people can come down and ask their questions and then the other thing that these have in common is that we're in the blockchain space we're building systems for adversarial environments but as humans who've had relatively nice lives it looks like everyone here has all their limbs you know things have worked out pretty well we focus on the carrots and not on the sticks right you can't you can't have a functional Byzantine fault-tolerant adversarial tolerant you know tolerant of adversaries type of environment and not punish people so you know we have to be as a unity prepared both to be punished and to punish others and you know I could get all philosophical about it but I won't I'll just say you know looking to operant conditioning looking to sort of like behavioral economics so that's actually another point that I'm gonna get to the questions very shortly but the point is that a lot of times we're looking at economic models that I think or economic you know the history of economics has not until very recently economists didn't really have a behavioral model a behavioral approach or a behavioral modeling like approach like a neuro economics or behavioral economics approach and so you should Oh in game theory suffers from this I mean game theory for the most part exists outside of this sort of behavioral lineage and so when we're designing our systems we need to be aware that we may be referencing literature or there may be even contemporaries producing literature today that may seem instructive or helpful but doesn't really map to the adversarial agent based sort of environment that we're really operating in and I think we have to be very sensitive to that which is apropos of me asking the audience questions so I'm gonna answer some of the questions from Twitter that we're just very kindly posted I'm not probably not gonna in tribute those you can just go to Twitter and attribute them yourself I'm not gonna delete the tweets or anything if you could call upon a higher power that has the ability to change one thing about aetherium what would the request be doesn't have to be technical that was a really good question I think I would I would have I don't know there's like an obvious snarky joke that I'm not gonna make but that the polite version of it is I would have had more history more computer science history more voting history were in France a history of democracy of democracy 200 years ago people started really thinking about democracies here I I've had more of that historic knowledge inform the development of aetherium that that's probably especially both on the governance side of the foundation itself in those early days of early formation as well at the.you know the technical designs as well next question how would you retire or not retire eath 1.0 when 2.

0 is live that's a little technical but the long the short version of it is you keep the old state available the old data available but you don't allow to be updated and you do the updates via an overlay on 2.0 so you have sort of this which is kind of wonky but you know you also have tooling that dissuades people from making updates to 1.0 in the first place but yeah I used to have to support some sort of update functionality but it's mostly like migrations how would you implement an etherium to inflation Treasury like I wouldn't really I mean I guess I would simulate one it's necessary so I I don't know it's like a question I'm not a big fan of which is what I said in Twitter but yeah my answer is basically I'd run a simulation there's not really an easy succinct answer to inflationary treasury management I I think that you need strong governance it's just it's just obviously impulse in tractable problem that trying to do it purely algorithmically I think there's an appeal to having a commitment via algorithms right sort of being like we can't do these things but to then be like you can't be super prescriptive because that doesn't really you know adversaries have an advantage and there's other issues with being too prescriptive I think that you should be a certain level of prescriptive if you're going to be minting well like obviously like I like you definitely have to have some rules for minting new tokens in the future but you also need to have some rules that are soft those are like programmatic rules but you'll need some social rules to address like what happens when that doesn't work taking all possible factors into account what do you think is better for the a better way forward for aetherium one focused on improving and optimizing a theorem one X and layer two tech to focus on aetherium 2.0 in the transition to it yes ethers trying to do both but I think this is an interesting discussion it is an interesting discussion which we probably won't get into here but yeah my my Wow so I worked on Casper a lot like in the early days you know we had these weekly calls and I was on the calls and active and we talked a bunch I'm very optimistic about CBC Casper I really liked about to talk yesterday so I think that that research should continue but if I had constrained resources my primary concern would be getting the thing that I've already deployed stable and working and like have a reasonable projection into the future right I mean right now if you if you do the projections or how the future is pretty bleak so you know people have to work towards actually built in a system that has some kind of future and and so I think for me personally I would be extremely focused on 1.0 I would or 1x really I think is a better way of thinking of it and I would be in particular I mean well I already do focus on the areas of data availability and how do you access old historical state because I think that there's some dishonesty around what is and is not necessary from the historical state I'll get into that later if I get all ranting about that but that was kind of last year so I'm not really trying to do that so much this year how much at risk is the etherium network from bgp attacks / firewalls which is a funny / what note density distribution do you think the network has to reach to avoid stalling apart from total internet shutdown how much of an improvement in band okay this is a bunch of questions how much have an improvement in bandwidth constraints can we expect with the theorem 2.

0 so the last one I have no idea I mean the theorem 2.0 is like a pond napkins right and no one knows what the bandwidth constraints of that are going to be how much at risk is the etherium well the network isn't really at risk from BGP attacks but your service is and your services at dire risk I mean you have to be extremely careful about BGP attacks one of my took like COD quasi mentors I mean he had laughed if he sees his SAT as a mentor but he he he wrote a lot about BGP attacks in 2007 they're a real thing people have lost money in the Bitcoin community due to BGP hacks I haven't gotten it confirmed from anyone even like a little birdie or whatever but when you read the reports from coinbase about the et Cie hack it looks like that was an eclipse attack which may or may not have I mean it was executed on the network layer in part so BGP is in the network layer so you need to be concerned about the network layer as a service and obviously if you think of the network the etherium network as a collection of aetherium services in that sense we all need to be concerned about BGP because if someone manipulates the network the the BGP the internet layer too much nothing nothing's gonna work but how much is that risk in practice it's very low because like you know Google cares more about getting hacked by the Chinese then like frankly etherion people care about that so they're gonna like protect the internet for us so that's why the risk for us is relatively low at this point I actually did go to ie TF 99 and discuss bgp security and I actually asked the people who you know deploy the Internet why BGP wasn't more secure and it was a very enlightening discussion but long story short the blockchain community has an opportunity especially at the IOT side of it to fix this sort of core Internet vulnerability because IOT devices that need to connect the block chains need a way of trusting the networks that they're connecting to and they don't have another way of doing it without a secure bgp protocol so I think it's super important I've done a lot of personal research into it not really so much from I asked perspective what no density distribution so so yeah I think the note density distribution question is interestingly I mean yeah we would like to have a good distribution across the various asns of the Internet to provide robustness we might also want microwave radios we might also want things like a Thea mesh we may also want satellite uplinks these other things to maintain robustness something like Bitcoin fiber to row to provide robustness I mean I think those are all nice to haves that are more important or not more important but facilitate a good node distribution so so yeah and I don't think I think total internet shut down is an interesting sort of hand-wavy concept because the United States government as far as I can tell has the autonomy to effectively completely shut down the internet for everyone and there are I think you know people who have thought about this and have made some countermeasures but ultimately you know if the NSA wasn't like a fetishist that just only watches we would be in a lot of trouble because they have the ability to shut down everything but luckily for us they just like to watch so they're not gonna do that oh what's the most likely way in which aetherium tube will fail and what would i do about it so I think the most likely way aetherium tube will fails it'll never happen that's just my sort of historical background in life like I just generally think when you're suggesting these large transitions there's a lot that can go wrong there's a lot that there's a lot of existing momentum there's existing miners is existing institutions there's already friction to change whatever the multibillion-dollar market cap is of aetherium those people have invested money they may not want to have aetherium 2.0 happen for a number of reasons so I think those are kind of the biggest risk just sort of normal failure just normal large-scale project collapse in terms of like technical risk I'm not a big fan of the beacon chain model so my basic critique of it is that economic load balancing mean is another way of saying the attacker always has a tempo against the network because the network has to advertise how it's going to load balance in advance of load balancing the attacker can the load balancing isn't doing anything to deter the attacker so it to load balance in an adversarial environment you really have to give the various counterparties in the system the ability to sort of form on their own trying to load balance under consensus in an adversarial environment doesn't really those death sentence doesn't make a lot of sense to me from you know based on my experiences I think that that so my proposal which I made you know sort of casually and passing to people over the years is app starting so you create a marketplace for validator so you have a you have a beacon chain you have a main chain and the main chain primarily what the main chain is is for is for bonding so I have proof-of-work value that's on the main chain and maybe I'm not running you know some huge number of proof-of-work nodes but we've sort of done that already we've already minted a bunch of aetherium so we don't need to keep minting more aetherium in that way so we have this main chain proof-of-work tokens before two counterparties get into an agreement we both bond or one party gets into you know bonds and that advertises to another party that they're serious they do a deal basically a smart contract agreement that is in a much different language than what we're using now on the main chain that includes the validators which are sort of like this third-party arbitrator and the two counterparties just to keep it simple so all three every about all three of us get together on the on this beacon chain we make an agreement we then go off chain and use proof of stake to actually execute on that agreement and that scales obviously very very well because very little is actually happening on the on the main chain it's also you know cleanly recursive right it's it's very obvious to see how you make a recursive version of that where you're you're you know you have deep trees of in sharding and a lot of complexity there it's actually very straightforward it's also obviously grossly under specified because my talk isn't about specifications for her scalable charting solutions so but that's sort of hand wavy the my suggestion as an alternative to the beacon chain and I as I said because the beacon chain I don't think will load balanced so that covers all the questions from Twitter a lot of people were said they're very excited to be here a number of them are here so it's very nice but as anyone in the audience have any questions I'm happy to field questions from the audience I suggest that you come down to where the projector is and we're also able to have just a discussion if there's no queue I mean I'm in Europe so it's a queue so if there's no queue of people I'll just have a conversation with Boris that's cool too we can handle it any number of ways good morning morning I am interested in your thoughts on using or adding dev rewards to aetherium as a way to fund improvements maintenance of the network that's a great question so I think that I think that there's a fundamental confusion or maybe just a lack of understanding when you know how would you know in advance it's not really a criticism of people just sort of it would have been nice if we'd known at the time aetherium was definitely under specified on the governance side and so I think that and I think tezo's is going in an interesting direction I tend to side more with the cosmos model of governance although I think that there's some interesting stuff that can come out of Tasos like it's hard to say you know what will will not come out of that and so that was sort of an obvious preamble I think it's a governance question right I think that if we had a functional governance system then the token holders the miners the existing doves DAP developers would be making staked you know signals they'd be making weighted signals you know you'd make a proposal and then every user you know and only like 5% of the users are even going to vote but like 5% of the users could then sort of signal slash vote on these proposals and we can sort of make decisions about that I think it's obviously a good idea I think it's obviously a way to actually run a sustainable system but I understand how you know we didn't get there in the first try I mean I think that that would be a great thing for the foundation to sort of spearhead it would be a great thing for consensus the foundation I mean how much does by Nantz rely on the sort of openness of etherium to exist right so I think that those major stakeholders if they could get together and suggest an inflation that would be awesome but you know history has told me that's not really how the world is going to work out I have a nice question over here as well so we've we've got a number of groups that are both collaborating and competing to produce what we might call a theorem 2 point X we've got the foundation itself we've got Kaspar various different caspere's we've got Definity we've got polka dot we've got the list gets longer and it's it's it's a substantial number and each of these groups are collaborating to some extent they depend on each other for information flows they're also competing for the value where the value migrates in the long term so my question is how do you think this collaboration competition is gonna play out that's also a really good question so I think probably the main problem in answering that question is that block chains aren't businesses right I mean that's really kind of counterintuitive you would want to apply sort of a business logic to it like this thing came out recently like Palantir has never been profitable I mean that's like mind-boggling like how how does that make any sense that Palantir this huge like industrial machine has never turned a profit right so and no and by no means a disrespectful claim to Definity but like I can imagine Definity becomes the Palantir of blockchains right there they have raised all this money and I don't understand how VCS work so somehow they've raised all this money yet have no obligation to ever be profitable and so they're just churning out like great BLS work great random beacon work great all this you know great wasum work and they're just like some sort of weird charity effectively that like twenty five million dollar token sale right so I I think that you know that sort of that sort of kind of hard to foresee consequent situation I think it's gonna happen to a lot of these chains where they're not gonna die right there's this zombie phenomena Simon formerly of Oh Joe a consensus did toke and curated markets thing he had a token and he was the only miner right he made his own token he was the only miner is running and someone was pumping his coin one day it was like oh that's going hot there's no activity on this project I'm the only activity I know there's none why is the price going up I think there's a lot of these sort of weird phenomena that are in the ecosystem that other eeeek that nothing prior to this has had so I think that a lot of code reuse is how we're gonna see a lot of collaboration is just people stealing code I see a lot of adversarial sort of open collaboration right I don't see like people talk about like acquisitions right like oh we buy a coin we merge it we do a one-to-one exchange or whatever like a like us like a stock like a sort of a company merger sort of concept I'm not so I'm sort of skeptical about that model but I think that people definitely steal continue to steal code I mean I think that's actually one of the best models of collaboration in the space not the collaborative code usage oh hey I got this BLS lobber a from Bram it's awesome but like the people are just straight stealing right like there's no there's a lot of that and I think that's like also a cultural thing without getting all weird about that I think that some people are culturally they just don't stealing is a cultural construct and so for them it's not really what we would think of as stealing and again I'm in Paris so you know I but I think that sort of dynamic is gonna continue and I and I I mean at this conference there's a web three track it's a whole track to web 3 I mean polka dot should buy any you know sort of conventional measure beat in direct competition to etherium I mean Tron I mean II OS has issued tokens on aetherium Tron has issued tokens on aetherium they're their initial token their initial sale was on aetherium those are competitive projects I mean directly competitive projects and and and Tron is a fork of aetherium J as far as I can tell oh I've screwed up the punch line to my joke but from a show of hands in the audience I'll tell the people watching at home how many people here know what aetherium J is okay that's like let's call that just under 50% how many of you actually like used it or try to run it or had any infrastructure I think one person like kind of almost half waved their hand and so and then as sort of this third question to be a little troll ish so how much do you think the etherium J codebase is worth right if you call it a theory of J it's no offense to Roman it's it's worth you know maybe maybe $100,000 or some ridiculously a million dollars or so less than the dev hours put into it but if you rebrand it as Tron and you take out dev p2p and you put in a bunch of G RPC calls all of a sudden it's like a billion dollar project like who knew right you know Roman hopefully is got a boat from those guys or something I mean they should do something for them but yeah I think that those sort of really counterintuitive dynamics are what's going to define the blockchain space moving forward and I think it will work both to inoculate against sort of hegemonic you know the Google's Facebook's apples of the world but I think it'll also you know slow down development in certain situations where you could have just got those five guys in a room and they could have just got it done in a week but instead we've you know been hemming and hawing transcontinental II for three years about you know some dumb thing that just wouldn't have been a problem in a normal institution and I believe there was another question yeah yeah okay I put this on the on Twitter but um no I missed it there was late as writing is bad - I'm sorry that's fine okay so I'm just gonna read right here it's about state state build-up do you believe in the medium term future maintenance will fail to sink dude a state build-up and how would such a crisis play out Oh I mean solvable yeah it it's absolutely resolvable shout out to let's see real ledger watch who's bending over backwards you know fighting dragons the guys Forge and swords then go fight the dragon he's mine in the or I mean he's doing all the work so you know shout out to him I I think that is a tractable problem simply because I've talked to him about it and he seems to believe it's tractable and he's done so much research you know he really backs up what he says with the research so I think it's possible it's a real issue I think you know good news bad news one day gas is gonna turn off the event log you know index that it provides and then you know good luck for the 90% of real apps that rely on the event log to do something so the work that I with with maker we're transitioning off of the event log for other reasons but you know that's sort of the side effect the knock-on side effect of that that's really helpful is if the event log were to go away our system still works which is I think you know it's really difficult to reason about how to build a robust full stack on the current aetherium node code base you know if you're relying on gas or parity it's not clear how you actually write your app and that's something that we work on a lot at vulcanized I'm not really going to talk about my own projects I don't want people to bug me but yeah in terms of the current configuration of the network how much time do you think we have again I mean Alexi white starts playing out somewhere on the Internet he has like a chart where he's like we're doomed Boris do you happen to know offhand what the number is two to two to three years sounds really good I just realized that my art archive node wasn't on didn't on the constant domo fork so I know that my archive node like I'm literally gonna have to pay a dev like thousands of dollars to like go in and do like surgery on our archive node to fix the six days you have at one time I knew exactly what the number was because we I was arguing with Nick Johnson about it but I've forgotten the number you have like way shorter time than you'd think for your no it's four hours I do remember the number if you're a full archive node is down for less than four hours it's dead and you have to rebuild it so when we're talking about these state issues I think that you know the weird the weird sort of situation that you're in and trying to archive all of the state today and then we're like I said if you're out down for four hours you're out of the game but at the same time to say that the growth is only gonna be like cataclysmic two years from now that's I think I mean again I trust that number of two years but that's to me in a sense very optimistic because I feel like I feel like I'm you know I don't need to I don't need to die to go to hell right I think I think I'm already there right I already have serious issues dealing with the state so two pieces there one is two years is a linear projection of uses of usage of aetherium so if we think more people are gonna build on the theorem and we don't give them better tools on how to build gaps ahead of that time then we likely have less than two years a Lexie's roadmap of how long it will take to do all of his state rent plan is three hard Forks which is roughly 27 months that's funny I love that that's I don't know you couldn't you couldn't write that in a movie right you wouldn't use a speeding train it's amazing yeah yeah I mean someone needs to give Alexi more money like you like or any money can I share a couple yes I wanted like this is these are important things I want to make sure everyone's on the same page so Alexei has received some grant money so yeah we're getting support nothing doesn't deserve more right and I agree he's like I envisioned him as kind of like Atlas like carrying the weight of the etherium world on his back which is kind of you eluded to him I know the number so this all started at Def Con and this was like Peter Sivaji who originally provided estimates tonight I can't remember what he said 18 months or two years or something but that was his estimate right for like how long we have and yeah like we're hoping that hard Forks happen every six months or so right so best-case scenario is like 18 months so like we're definitely coming down to the wire a bit but just to be very clear it's not like aetherium switches off if we don't get there it's just that like the node count starts to decline you know so it's yeah well you and I are have a different level of cynicism in our delivery so we can we can play good-cop bad-cop if you want well no I just I mean it's sort of like you know that passage of time is the curtain being pulled back and like surprise its centralised right I mean that's really because if the network continues to run during that crisis and updates aren't happening you're getting rid of you know there is no illusion that you're in your home running an ode for one up for my dev team I had to just go out to like this place called micro Center just where you buy computer hardware and I it's been like over twenty five hundred dollars on hardware to build to make a relatively powerful macbook able to fully sync an archive node right I mean that's not running aetherium on your phone right there's there's like there's like a divergence and the narrative there like it's not you know three thousand dollar investment is nothing for a professional miner but you know and I'm maybe hopefully the only time I mentioned this dude this entity's name satoshis paper sort of assumed that it was a p2p Network and that there weren't these this hierarchy of miners and users right so if you're gonna use these sort of proof-of-work models these sort of Satoshi you know these Nakamoto ask models you you do need to have this broad distribution of validators of course there's other models that I think are legitimate and valid that don't have those distributions so I mean but again we kind of have to have this self conscious self-aware discussion about it we can't just say like oh we'll be fine someone's gonna show up and save us like nobody's coming to save us is there a possibility to create a IP where it reduces the gap between an archive node and a full note because deprecating archive nodes would be great and I think after a few years we kind of know what it term is all about and for example token transfers are 19% of the smart contract calls and it's about what 60% of all transactions is is there not a possibility to remove like the necessity of tracing and archiving all of these events into what's something that could be built into what the full note does that's a really good question so I now I do have to pitch what I'm working on so that is what I'm working on is building a system where you can tell the Gath client what you want to archive and it only archives what you're interested in and this is very similar to how Bitcoin actually works with accounts right Bitcoin tracks your account and keeps the table for you and I forget the names all the different tables and Bitcoin but it keeps track of the table for your balances so that you know what your balance is but it doesn't keep the table for every account in every and all of Bitcoin and in aetherium is lacking an analogous feature and so that's basically what we're working on it vulcanized dB well it's a semantic discussion right I mean that's we're we're we're submitting patches to death and parody like it's a you know again a series of patches that would facilitate that so you could go into parody or gap and say I want to watch this contract address I want to watch this event topic give me you know spit out the trace just for that so what's in there and I'm you know I'm pretty cryptic so I'm not gonna say you know what people are actually submitting those patches but I'm paying them so that's that's how that's happening but yeah so I and I think that moving forward that is the answer is this sort of we used to call it it's a very confusing name but when I was with more working on tinder meant stuff we called it like a light node as opposed to a light client right it's not it's not just like getting SPV proofs but it's tracking all of consensus but it's not tracking all of the history necessarily right it can't necessarily rebuild everything and I think that's just an architectural thing that'll emerge I mean I guess this is me announcing that to the world or whatever but I think that's the way to do it and I think that yeah the archive nodes as they exist today again no disrespect to the to the developers who who design the protocol as it is it's not sufficient it doesn't it doesn't really solve anyone's problems the block explorers don't really use it and as a normal user there's no reason to use it and so it's not really serving anyone and it does need to be modified I'm interested in view and like Grant's distribution there is like several entities in his space that are like giving out money to like different projects some of them like may hop may have like different focus do you think like the you know like the money given to Grant's is like spend effectively do you see like a huge difference between the different entities of working days I can't say much about the spend I mean I can say about what we know about the rewards publicly because I don't have any insider information about I mean we apply to Grant's we have I was ordered in theory and Foundation grant frankly I think I asked for literally I shoulda added a zero to what I asked for I just like was off by an order of magnitude I didn't that's a whole nother story but do I think that the allocations as announced are efficient no not really and I don't think that aetherium is the worst offender in the space by at all by any means in that in that regard with and that's all I'm gonna say about that within the etherium ecosystem itself I think there's a lot of opportunity to optimize I think there's a lot of opportunity to coordinate I think there's a lot of opportunity to say to people like we're gonna give you this money in the condition is that you participate in governance right I mean that that's like almost an obvious it's almost like a no-brainer I feel like a lot of people I don't mean it in like unnecessarily a derogatory way but when you're running a business sometimes the parsimonious thing for your business to do is to take the money and run right and it's a grant and so a grant doesn't have a lot of strings attached normally and I think that aetherium I wouldn't say has been victim to but they haven't necessarily of the foundation hasn't necessarily aligned themselves in their issuance of grants with rational business behavior hopefully that wasn't too polite to be obtuse but yeah I think that they could be done it could be done better in the sense that there could be better coordination and it could still be a very loose like decentralized coordination you know you have to type in this chat once a week when we vote I mean it could be very minimal but I think already we're starting to see an inconsistency a non-linearity in what contribution people are going to provide what they're claiming they're going to do and how much money they get they get announced because we don't really know if they get that money or not because the announcements don't really map to your payment schedule I can say that from my experience so maybe those maybe those teams actually have very reasonable conditions and they haven't been actually haven't actually received the funds because the funds aren't distributed on chain or anything we don't really know right but the opacity you know that's a too opaque of a process and from the outside it just looks like it's not very efficient hopefully this is not a controversial question what are your thoughts on polka-dot technically as well as just the ecosystem I'm gonna start at the beginning I don't know how they got latex to print on a polka dot background for their white paper but they ensured that I would never read it in its entirety by doing that so I'm gonna start at the beginning I've never read that whole white paper because it's on polka dot paper moment of Zen I don't know we should all reflect on that that's some sort of Zen koan you're gonna open up the coffin there's only gonna be a shoe Regis and cones folks I there's some sort of deep brilliance there that I don't understand I'm very skeptical of fishermen as a concept as ek snark fishermen again you know all these all these little groups that were in are also deep right I'm sure there's somewhere in the universe some Zeke a snark construction that will work in this in this in the hypothetical that they're describing but as I understand the technology today fishermen just won't work and they won't work in the general case so let me see if I can give with my limited knowledge of poccadot let me see if I can articulate my position without being too much of an idiot so it's unclear to me so if you have a system like polka dot or cosmos they're I think they're very similar and that's not a coincidence right well I don't know quite how you'd label it you have to make certain strong claims about the relationship between the hub chain or whatever you call that chain and the the side chains or whatever the I do parent and child but it's confusing because the parents are the anyway so you have the hub chain and the other chains that are connecting to it and the hub chain has to be the fastest chain out of all of them if it's providing a union of state over all of them and so that also means that in the case of polka dot if if the fisher model actually made sense as i again my understanding someone from polka dot could show up and be like Rick we have Paige blah blah blah page seven paragraph three says you're wrong okay fine but as my from my understanding of it the fisher process so let's say that you have like a three second block time that would mean that you would have to build the fraud the fraud the the the proof improving that the block was invalid that was published from the pair of chain was invalid you have to you'd have to build that proof distribute that proof and all the evidence for that proof within less than three seconds because then you have to vote on if that proof was legitimate so you have to get the block all that all of the polka dot validators would have to get the block from this other chain they have to get the client the fraudulent claim there to get all the evidence for supporting the fraudulent claim which is at least another block so at least two blocks you have to get all that you have to gossip that around everybody you have to construct this proof and then you have to now run your vote so I don't know about all that again you know well ironically right if you if you're IBM and you have some sort of you know alien tech from Roswell computer where you can do all this centralized and in time you don't have any problem but to claim that you're going to do that in a in a permissionless network i I just no I don't think that's gonna happen we have a question from Ustream chat box do you believe that lowering the block Gastelum it could be a solution to prolong the life of earth 1x lowering the way what oh I get it I broke my leg so I'm gonna walk slower is what that sounds like to me I mean maybe just like go to the hospital and get a cast like maybe do that instead like it I mean it's a very engineering it's a very again it's actually it's a very parsimonious solution to the problem um we'll just make the blocks objectively shittier so we have more time to fix the network okay I you know I mean there's worse solutions than that frankly but but yeah it's not an optimal solution I'd rather not do that we've another one from Twitter if that's okay oh yeah yeah someone I mean thanks I'm glad we've emergent an emergent process here this is wonderful that's let's keep in common how hard will it be in your opinion to implement state rent on aetherium that's a good question state which state rent or storage management fees whatever you want to call it yeah yeah you don't want rent you don't want tax you want to call it some I'm so not a marketing person I think easy again Alexi gave me a hole he had a whole breakdown for how to do this again very impressive so I you know I kind of would just say to whomever that is go find a Lexi's work and read about it my personal view is that it almost doesn't matter how hard it is since it's imperative it's existential so however hard it is it doesn't matter you have to solve it or whatever bitter pill the community has to swallow they have to swallow it or they're not a community anymore so in that sense I guess in that sense to be to be truly strict about it that the questions are relevant it doesn't matter how hard it is but I don't think from an engineering well let's let's all have an honest moment here I mean how much legacy aetherium code is there really like makers under hyper active like development right I mean what are people really doing where they have a legacy code that can't be programmatically updated if we have a rent situation I mean because I think I think for example ERC 20s it'd be very easy to say okay well some of your fee goes to the rent on your cuz you know your each account has its own row in the ERC 20 basically and if you get your you know your token then you have to pay your rent on your token I mean how hard would it be to almost like at the minor layer within the actual node software enforce that rent in the ERC 20 context right that'd cover again 85 90 percent of the traffic excuse me so you know that's kind of ugly from an aesthetic perspective you know that's not a great sign of the community hey everybody we passed your C 20 is whether you liked it or not deal with it but you know again it solves the problem you've mentioned governance several times he said cool word like voting would you like to give a parsimonious answer around aetherium governance whoo so we're having flashbacks to when I was in the other smaller room I really think that people should look at a lot more voting Theory political science like there's you know arrows and like all these people have like done all this research over literally hundreds of years on voting so we might want to look at that stuff from a governance perspective I think it's also worth noting that governance is something that you know I'm gonna say it's uniquely human but it's something that's existed you know literally pre humans like whatever animal pre-k's you know predates the human like they had leaders and governance right governance is like a really old thing so I don't know that we necessarily need to invent a whole new language and a whole new concept and a whole new model um I think we would be better off again there's so much money in the crypto space just go to a university and just like buy a reputable department and like get them to solve the problem and then just hire some devs and just implement it and just move on uh you know it's so yeah I but I don't think anything that's gonna happen I think that we'll continue to have the deadlock that we have in aetherium for as long as possible maybe longer than possible I think I think the governance failure may may be the failure of aetherium I think that there's a very high likelihood that the stakeholders don't come to alignment you know by Nancy issuing their own coins on their own network you know how much how much do they need to align with what aetherium is doing and stos in general which I'm I work on some stl projects I'm generally a fan of them conceptually but you know aetherium doesn't aetherium has I think a real user base in a real use case which is the sort of a legal slash extra legal Financial Services but that's not a world computer so maybe maybe we don't promote that because it's not a world computer all right sorry yes I think you just touched upon it a bit and assuming that I have this brilliant app idea and the money to pay you as a consultant what would you recommend me to go for in terms of protocol and I think you already mentioned your opinion about polka dot maybe more focus now on aetherium and cosmos and maybe more specific for what type of use cases you would go for one or the other yeah that's a that's a good question there's some free a bunch of people getting a bunch of free advice from me that's great I'm happy to do it the generally what I advise my clients to do is a federated proof of stake sort of side chain on aetherium so use tender mint do your internal agreement with tender mint have your validator set be represented in a etherium on chain multi-sig because etherium is available today and you can actually sort of rinse and repeat that you can have your you know seven validators in private or however many validators i mean their permission they're not really private like there ought the whole point is that they're publishing their I don't like the term private blockchain as an aside because the purpose of a blockchain is to create an auditable transaction log so it's necessarily not private a private blockchain is a oxymoron it doesn't really make sense so you know you take your your permission set and you can synchronize that set of validators across multiple chains so you could have that validator set synchronized on aetherium synchronized on Tron synchronized on you know on a OS you know II else doesn't have finality so I feel very uncomfortable talking about es at least Tron has a finality right so you can synchronize on on one of these you can set to synchronize your validate or set on these other networks do your actual business in private on your private network yeah on your like physical physically private network and then publish your proofs to your counterparties on these public chains and so that's what almost every project I work on has that structure to it starting I started doing these in May 2016 and I've worked on I don't keep a perfect dead reckoning account but maybe like eight of them and they've all sort of used that same architecture some of them decided to invent plasma instead instead of using tinder mint and others have used other things instead of tin Dermott but most of the I mean we start with tender mint because the cosmos SDK is super good and works right now and you can build stuff with it right now so that's my general recommendation I think in terms of there's some details oh don't run your own virtual machine don't run the EVM don't write a new VM try to avoid vm's I actually like TX VM I was reading about that recently from the Staller team so or I'm sorry if I'm miss attributing the team but Dan Robinson was an author on that as well as this Olaf guy I like I like that VM a lot I I like the vm's that are coming out of Casper labs so they I feel as though they're working on more than one they're very similar the TX vm + that VM are very similar packed from Kadena Kadena I forget which way you're supposed to pronounce it they have a good VM but you know for the most but generally I say for most people you don't need a VM you can just it's probably easier still so just very carefully write code and whatever language are comfortable in because you're not you don't want people publishing arbitrary contracts anyway you want people to publish you know another loan you know you specify what a loan is in your system and okay we're issuing another loan you're not like okay random person issue something that's not alone it's like it's a network for issuing loans so you don't need to really have a VM in there not a smart contracting language at least so that's let's see you is that the key points to what I pitched to my clients a high block fast block times so fast block times a good story around checkpointing and how to handle empty blocks because in these systems you have mostly empty blocks and you don't want to be spamming everything with you know a lot of overhead on empty blocks and then we also do a lot of work to make sure that in these federated systems that I work on almost all the work that I do is actually in making sure that there's an incentive alignment between the parties that are transacting in this multi-sided marketplace which is what which is what these chains are is there any more are there any more questions cuz I'm just gonna spend the last couple minutes yeah go for it maybe a quick one um one of the challenges we're seeing and this is closely with it related to governance in making any changes including if your iam 1x including state rent is that there seems to be like a social contract with developers right so this is actually why Constantinople was postponed even something as seemingly benign is like lowering the gas cost like pisses people off because they were assuming these invariants that existed that maybe we I don't know we never had this conversation so the question is do we have an implicit or explicit social contract with developers and if so what is it and what do we do about it I don't have time to answer that so I I think that there is both an implicit and explicit contract with developers and I think that the etherium community so if you think about the Linux community which I'm familiar with like the actual Linux kernel community it's very clear when you go on to linux dev who's talking to whom you know there's hardware manufacturers there's people who are trying to get hardware to work and they're not the manufacturer and there's people running large amounts of infra and everybody knows their right having everyone in the Linux kernel group knows how to write a kernel but I mean even though it's a very you know they'll they'll teach you how but like people people are in that milieu right they understand what's going on one of the problems I think etherium has from a messaging perspective is that is that uh you know we talked about building consumer apps consumer adapts and I think that that is not I don't think we're far enough ahead to do that or I don't think we're developed enough to do that which is going to segue into my last bit so I think there is an implicit agreement and an explicit agreement with developers and I and I don't think that that's between all the members of the community not just between the foundation and the some other group it's everyone collectively has this Agreement collectively and so what so I'm gonna I'm gonna part with this so I did this in Reverse I gave a talk where I did the question-and-answer first and then I'm gonna give the talk part because it's very brief it's called the deep stack problem so I'm gonna leave people with this with this problem in the blockchain community at large we have a group of people that are claiming that they'll be able to create a new Internet in a relatively short period of time and they're making this claim without really having like a company they don't have a group of people that you can point to and say these people are gonna make this thing so we don't have in the blockchain community at large there are people launching satellites right that's that's a technically complicated endeavor managing satellite communications we have people building chips we have people manufacturing Asics we have people talking about Asics supply chains that is also obviously a deep domain of expertise where people need a lot of ability we have enormous conversations about governance right again hundreds of years of research in governance we have people designing their own you know not invented here designing their own you know smart contract language is you know programming languages huge domain right distributed systems huge domain consensus algorithms huge domain UX but what is react like for I'm I've been coding for a long time when I look at react I'm like this isn't javascript what is this I don't even know what it is so we have reactives we have normal devs we have ASIC manufacturers we have satellite launchers and and we have a bunch of PCs and other I mean another interesting about the blockchain community is we have investors that are of a novel type that you know crypto crypto investors a double entendre right so we have these we have VCS we have crypto investors and and they are all saying we'll have this done in like seven years and I'm like dude did you use Usenet do you know it was a decade before Usenet happened and Usenet was always trash like from a user experience perspective Usenet was the bomb in 1990 it was like blowing up houses it was the bomb and it was still trash you still had to sit and wait forever to get your news your news group to update and that was the state of the art and that took a bunch of nerds who were all experts in in military research labs decades to make Usenet so the idea that I love this ragtag group of people but we're a ragtag group of people we're gonna show up and we're gonna like you know sit down unzip our pants and just like code up a whole new internet including networking satellites UX Hardware ya know that's not gonna happen guys sorry we're not that plan is not gonna succeed I left off data availability problems I mean I just even in this example I'm being very tourism I'm not really enumerated every single issue every single deep research topic that needs to be thoroughly understood before we can fulfill some of these very simple claims like aetherium is a world computer so I just want to I'd like I think the biggest problem like within the conversation like these different parties from investors developers and everything and even with the state rent issue or with the grants issue is that we don't have a clear vision we are a very decentralized community and the problem is that nobody wants to be highly opinionated about what the mission is so we need to be more explicit about the mission that would make it so much easier to define what these social contracts are so if the mission is to be as progressive as possible the state rent shouldn't be an issue but there's this concept of like oh you agree that this would be deterministically like this and now you have to just do what you promised but no the promise was to be progressive the promise wasn't to be like you're gonna deploy your crypto kitty and it's gonna look the same in 20 years like the premise of a tea room is to just continue developed in my opinion and we just need to be more explicit about that it just makes the social contract much easier yes and I'm out of time so all I can really say to that is yes but yeah thank you everyone it's been a pleasure [Applause]