2021: An Ethereum Odyssey

Transcript

[Applause] [Music] right mic check okay well thank you everyone for attending this keynote and thank you uh to the crypto valley association for having me today i was asked to present something about ethereum also why can you talk about ethereum um it's been two amazing years of market recovery since the previous bear market so i expect our industry to continue to grow a lot and also to welcome many new talents so this talk is about um well

remembering where we came from and uh seeing where we'll be going further um quick piece about myself my name is jerome i run a startup company called comet.io we do gaming and a couple of tooling for d5 you can join our newsletter and obtain some nfts if you if you'd like you can email me there i'm also a professor in economics at clam and engineering school and i'm the lucky president of ethereum france an association promoting ethereum in french-speaking countries i know many of our goers at the hcc conference are from switzerland so if you heard about this conference do think about coming there it's a cool

conference not as cool as one having one in zug the crypto mecca but still and it will happen again in uh in paris in july 2022. so a little bit of prologue do you remember do you do you remember those days october 20 2008 uh probably not not many people were around at this time who were interested in bitcoin but still this is the founding moment for our industry having an animus person or an animus group called satoshi nakamoto saying hey i get i just published this what do you think about that it turns out on this cryptography mailing list with many academics this caught the interest of many researchers of many developers of

many implementers that wanted to see this happen and the following months is january in 2009 we had the first version of bitcoin implemented and the famous genesis block being created where you see the the text written in the block uh making an appeal to uh a previously published uh headlines of the times chancellor on the brink of second bailout for banks flash forwards one year after this uh may 2022 if you like pizza and you like bitcoin you should remember these dates because this is apparently the first recorded transaction for a real good in the real economy paid with bitcoin bitcoin talk forum people were chatting about hey i want to

buy something with my bitcoin would anyone send me some pizza against bitcoin and they settle the transaction otc well you would go otc for 10 000 bitcoin they did otc on the forum and someone paid 2000 bitcoin for two papa joe's pizza flash forward a couple of months after uh the japanese trading card company mtv gox started to authorize bitcoin payment and bitcoin trading and of course the opening of silk road uh almost six months after this is where the the the tricky reputation of crypto may have started where we saw the um the race of having access to uh illegal products through dark web paid

with bitcoin even if two years after this this was closed um this this stain on the reputation of crypto stayed for a long time we still hear people that remember those days saying well bitcoin is for terrorists prison is for buying drugs on the internet bitcoin is bad and blah blah blah yet it helped us get headlines get get attention on what was going on in our industry in december 2013 i i have to tell you about this date of course this is where um vitalik betering published the white paper of ethereum and a couple of months after uh we had the crash uh

small crash of bitcoin compared to crashes that we have experienced after this but empty goggs uh closed went for bankruptcy after a hack uh the trial is still ongoing today that's a recurring joke a recurring topic you can chat with different all-time bitcoiners in july from july to august uh interim raised money through an ico through people committing if committing bitcoin to get ever and finance this crazy project that was ethereum the crazy project was building a computer on top of a blockchain emulating the a computer through a blockchain july the 30th the eternal mainnet went live

um there is not something fancy in the genesis block of ethereum it's actually a reference to the block on the test net where it was created no not too not too much not too much fanciness indeed but still i'm marking a nice day and another big date i think for our community and and um and the global industry in the global the global production industry uh was the hack of the dao since it um it was a live action of a governance action uh being taken place in the blockchain but i'll tell you a little bit more about about this so on the pictures you see uh marka palest the former ceo of empty glocks apologizing to the japanese authority a little bit like how carlos gone will go

in front of the of the of the authority after he was convicted by the authorities and you'll see also the picture of different founding members or different core developers of ethereum back in the early days but let's put things in perspective term-wise shall we so if you are recruiting someone that can do or claim to be able to do ethereum stuff or claim to be super proficient in solidity um when i try to recruit those kind of people i show them this chart without the without the the explanation and i asked them to tell me what it means what is the history behind this chart well i think you can see the whole

history about ethereum through this chart this is the gas limit chart uh the gas limit is how we measure how blocks are how big our blocks on ethereum the bigger the gas limit the more you can put in those in those blocks and in here you see it starts on july 2015 of course you have it changed in early early 2016 with the homestead protocol fork where the foundation said hey now the network has been running for six months is uh it looks fine like we should make bigger blocks uh you have the dow attack which i will tell you a little bit more after this and you have the denial of service attack or the shanghai attack right after this

and then you have the ico craze many people went into this industry because they heard about what about icos the fact that you can just buy a token out of the white paper and get a token that is almost liquid maybe the token will be listed and you can specul speculate a lot or commit long term on a specific product it's it it's basically a very nice really nice way to uh to make quick money but also to make quick financing of your of your projects and from many aspects the the ico craze was one of the reason we had uh the success with decentralized finance because right after the ico craze we went through hard months in terms of markets market

status which i think led many developers to think about how they can render those assets more liquid how they can how we can or we can um exchange tokens from one another now that we have them at hand so there is a good a good uh good thing to learn about this now that we have the nft craze uh we will we are likely to get a lot of uh application to be able to trade those nfts in the coming days not only the open c and variable and so on the block size had to uh had to grow also during the cryptokitties craze the first big nft project that went out and the last uh big push in term of uh what the the next pushes in terms of

block size were improvement in how the the interim client behave and the last big push was the eap 1159 1559 was it the typo here but i'll tell you more about this in the next slides so coming back to the dow attack who here remembered the dao well not everybody that's uh that's a pity so back in 2016 the ethereum community uh were having successes the price for going up people were getting interested the narrative was good and uh while while some companies uh some official companies like consensus were heavily financing the space and uh and and

creating startups and disclaimer i've worked at consensus before i love those guys some people in the community thought that we should build our own venture capital firm uh the vc were not so much interested in ethereum back in the days so what about creating vedau a vc fund but built on a blockchain everything was running in smart contract so i added a picture of not one of the developer of the dao but one of the developer of the apollo project margaret hamilton and the code of the dao from a solidity perspective was almost that big like how 600 600 lines or so of of code and of course it led to some uh

to some problem to some bugs so back in those days you could purchase a doubt again so 100 delta can against any fur if i remember correctly and the the project gathered about 150 million dollar or about 10 percent of the if supply back in those days for financing this so you can imagine the community being ready to commit 10 percent of the supply onto a vc project creating a fund for the community this experiment was short-lived of course because there was an exploit in the contract an unexpected behavior of how the ethereum virtual machine was behaving and it allows an attacker to siphon ether out of the contract of course the

siphoning was not a super straightforward as uh the the the different hack that we've seen in the default space uh recently uh you had to siphon the contract safe on the contract and put this siphon money into a new contract that would you would only get access maybe three weeks or four weeks after so it led to huge movement of effort out of the contract into new contracts uh when a whitehead group of hackers understood the behavior of the hack they tried to minimize the effect by doing the hack as on their on their hand and on the in order to protect the funds over other investors and it led to a fork on the july 20th that corrected this bug by changing how

the contract behave and giving back the money to all the to all the the participants this fork was not contentious uh like record of how much of the hash power committed to change the to change the blockchain was about 97 percent and for many critics it was the end of the hrm project because suddenly code was not low governance can be applied core devs and community members can ask for a change and this change can be sent to uh to miners that would decide if or not they should follow this this chain um i'm not going to tell you more about the dao if you'd like to talk about those good times please find me about

tell me about this about this talk but the sec soon after the the debacle i published a note saying oh this was definitely a security i'm glad you didn't do that another event was the shanghai attack um oh there's a typo here it was not in july but in uh in in september so during defcon 2 in shanghai it appeared that one of the fundamental operation of the interim virtual machine was not pricely priced well if you want to use a smart contract on ethereum it triggers operations like multiply store exponential read the memory here and so on and one of the challenges of creating a a computer on the blockchain was to

properly price those operations for having a good psjugo system and some of those opcodes were mispriced it wasn't obvious initially in the code until you discovered that they are mispriced and one attacker was spamming transactions that were calling those mispriced up code miss price mispriced in what sense well you pay some gas to run those up codes you convert those gas into effort to reimburse the miners that will execute those transactions but for the upgrades to be priced well it has to represent exactly how much of a computing power you need to use to run those up codes well what happened is

many nodes crashed the network shrinked very quickly under this attack because the the at the state of the client back in the days was not that optimized and not really able to handle those kind of uh of rush of consumption of resources and the reaction from the miners was to reduce the block size in order to cope for this attack that's why you see this uh big downfall in terms of of block size so in order for the network to be able to cope better against the attack the miners decided to uh to reduce the block size meanwhile many developers of the ethereum foundation and associated to the atrium foundation were trying to find a fix in

order for the for the network to still be able to live not completely broke down and also for the conference to start because it was uh 8 30 a.m in shanghai these days and people were like oh damn the network is going down what's going on should we do a conference or not in the end the network didn't stop it slowed and we managed to have the conference with a 15 minutes late 15 minutes later so here's a here's a remark to uh what people like the ceo of solana lab was saying when the blockchain went down for 17 hours no it didn't go down for 17 hours as did the germa as did the ethereum network during the dos attack of shanghai it

just slowed we handled less transaction but it didn't stop so having a blockchain deadstop is a problem of course having a transaction that rollback is a problem having a transact a blockchain that didn't give you finality is a problem but all of those problems are very difficult problem to solve and very difficult problem to handle let's continue to put things in perspective another interesting chart is the average block time so how long do you have to wait between blocks in the early days of ethereum the block times were about 17 seconds after the homestead fork we went down to 14 and now we are closer to 13 also

you may see during during big spikes here are attacks or big event in the early days and the huge race you have in july 2016 uh january 2019 january 20 and probably december 21 are built-in features of the protocol so that the protocol slows down if we are late on implementing proof-of-stake yep it's a feature called the difficulty bomb that i expect many people to not be fully aware about but it's something that was implemented very early on in um in the interim protocol because when you read the white paper it's clearly stated

we should build a good proof of stake system for our blockchain of course proof of work is the better the better the best um the best consensus mechanism that we know of at the time of waiting the white paper but we know that proof of stake can we think that proof of stake can be really good and we are going to build the good proof of stake system so behind this ambition was probably uh too much enthusiasm too much uh not enough realization of how hard it would be to actually construct a good proof-of-stake system and it took a long time to actually find the right based on the design rationale that each arm foundation and many contributors had

if you'd like to read more about this i highly encourage you to check out history of casper from vlad zamfir or exploring interim 2.0 design goals from ben eddington and so this is the this is the merge pull request where they decided to say yes if we miss the deadline of july 2017 the blockchain should slow down to to punish us for not being able to implement proof-of-stake and this has been a recurring gag because you see here every time this difficulty bomb kicks in uh what happened is that we decide to implement an eip that push back the difficulty bomb so we don't have to uh to care too much and we reset the we

reset the timer until next time i expect us to have to defuse the difficulty bond again in december but that's going to be the last and there is a major consensus among all the all the core devs of each room too that december will be the last time we have to defuse this bomb but meanwhile the evm franchise grew a lot you see here companies or blockchains that have built on top of the evm franchise evm meaning the ethereum virtual machine so creating your own blockchain with your own consensus with your own token with your own type of security of their own type of approach but building that on top of

the ethereum an interim basis ethereum basis being the interim virtual machine because you know that the one of the power of the interim protocol and the term community is creating standards creating codes that can be reused creating libraries that can be used to build better website better application and all talking to literature virtual machine at the end and on the right hand side of the of the of the slide you see projects that have tried uh sometimes successfully sometimes less successfully to build their own uh blockchain with their own evm with their own virtual machine actually and starting to build bridges on how your virtual machine should interpret

what comes out of the ethereum virtual machine because the community in the industry is so big using this evm so let's say polkadot with moonbeam you can run evm code in polkadot based blockchain thanks to last big effect last big thing that we've seen in the interim history is the eap 1555 it is is a conceptual change in how the transaction works if you have been doing lots of transaction lately you may have realized on metamask that you don't really choose gas price anymore now you choose a max gas price which is a tip you choose this and you choose that

not wanting to go too deep into how this works but what you should understand from 1555 is that now you are only tipping the miner you choose how much you want to tip the miner and the the the gas price that you should pay is uh expected to be at a certain level from one block to another so we have good predictability on how much the gas price should be from one block to another this is a huge change for a wallet provider for wallet developers because you can suggest a gas price with quite good certainty that the transaction will be included in the next block that your user will not sit there waiting and look at eaterscan and try to understand what's going on and expecting

the transaction to go in but you've been waiting for an hour and so on and so forth now what's good for the protocol itself is that whenever we have a huge volatility in terms of a gas price like if suddenly there is a peak or surge of activity of demand of being in a in a block space like huge demand for for transactions um then what happened is that some if are burned the fees that we are supposed to send to the miners before are now burned which is reducing the supply in in this graph you see in green how much ether are created to reward the person that participate to the proof of stake of ethereum in blue you see how much effort is

created to reward the miners of ethereum creating a block on the proof-of-work chain and in red you see how much ether are burnt every day because of this volatility of gas and of course having new collections new nft collections dropping every now and then is really pushing the gas price to lots of volatility so burning lots of if and two days ago was the first full day of deflation of reduction of the of the supply of if i said we have interim creation effort creation to reward the

to reward the stakers because ethereum 2 is now it has been now since uh the first december of 2020 a huge date for the for the ethereum ecosystem because it was the launch of what we call the beacon chain i'll get on to that in the next slides but it was also a huge date for me because this is the date my daughter was born so i had to run to the hospital thinking but well will my node kick in with everything we'll go well let's okay let's go to the hospital and have this child delivered funny story about this uh this genesis block is that there is a there is something in this in here called mr f was here as a as as um an interim two very detailed you can put a graffiti

uh you can announce the birth of your child you can announce your marriage you can announce that you love your mother you can write anything you want in the graffiti and this guy was drawn to be the guy to do the first to be the first to make the junior's block to do the first block and he put mr f was here because he apparently got a an autograph from vitalik buterin at a conference saying i'll just say mr hef yeah mr fahrenheit what is it jerome too by the way um difficult to put that in one slide so taking this challenge you can think of it as an airbrush so imagine a wooden airbrush the wooden part is the beacon chain is

it's a coordination layer and every range of hair here's here are blockchain of their own or like we call them the shard so instead of having one blockchain you have many blockchains or shards that are all connected through a beacon chain that coordinate those shards and make sure that you have proper communication between those shards um ideally all of those shards will have an execution layer so you'll be able to execute directly in the shard it will come as time go but the next step will be to take the existing proof-of-work chain and make it as the first child what happened to proof-of-work after this merge well

it ends we switch to a proof-of-stake model here's the program for the rollout we've done the beacon chain we've done the genesis block we are now going to the shard chains where you'll be able to have cross shot communication you'll be able to have the first the first the first chart actually you'll be able also to move the effort that you have at stake from the beacon chain to the the initial shot and eventually to the next chart um ideally it would be happening in december it should be happening in december because we have the difficulty bone kicking in and here you see a tweet from uh yesterday of one of the core dev of ethereum ii

client prism saying okay we have merged to the next stage of the each room to beacon chain status so the next step now is the merge i'm expecting this to happen soon and the phase two will be this so-called execution layer where you get a much better virtual machine to execute your contract and we have a better flexibility on the coordination between different jobs so you should be staking either through your exchange or for your own computer um today well yesterday the day of the day before yesterday we had about 50 000 51 000 unique depositors 51 000 unique people that

sent a 32 ether in order to be a staker so total validators of 250 000 creators of blocks when you are validating you are getting a return in terms of home or in terms of if so you can expect to have a 5.4 percent on all the if that you have at stake and if you look at all the effort that are securing the network it's about 33.2 billion but that's the price of the days before yesterday so it's probably a little bit better now and you should try to run your own node i know in this room we have many providers of staking services or many

exchanges that propose that you that's that you can stick with them but it's not that difficult to run your own node actually it's meant for uh the layman or the general computers to be able to run this beacon chain node you can go down to raspberry pi 4. the degenerational is that any effort holders with sufficient understanding of how to run a program on a computer should be able to run this on their laptop next step the merge when we will be at the merge there won't be any proof of work anymore like the the block of the merge there won't be proof of stake anymore i repeat there won't be proof of stake

anymore the chain proper frog chain that we know today will be secured by proof of stake by inputs from the beacon chain so we will be able to go out and say yay each room is green and stick if will be reservable to the ethereum one chain soon after and the current target is q1 and the the worst scenario will be q2 so here are some predictions to close this talk i think the merge will happen on the 14th of february 2022 because we all love ethereum and what would what should happen after

this merge is that the the sticking rewards uh should become a base apy for every d5 product that use if because why why would you put if as liquidity providers on an automatic making market maker like unisop or others if you can put them at stake this will probably be a huge opportunity for liquid staking providers like rocket pool stack wise lido and others um and if is very likely to become massively deflationary um as you see on the on the previous slides here uh if we consider that we are not going to oh my god i should have i should have put these slices after this one if you look at these slides and we consider

that there won't be a huge spike in terms of number of validators so the more validators means that there are more rewards given to the validators but in a decreasing manner so if the number of validators double we don't double the rewards we do quite less than that so taking for four hypotheses that we are going to have quite the same number of validators uh we are going also to have uh quite the same level of activity uh if we look at the current burn rate and the current emission rates of ethereum per week that means that we have about minus two percent of inflation every year so that's and that's just on the on the base the baseline scenario

uh so i think the this conference was uh opened as uh i was open with this uh with these headlines by bitcoin not because it will go to 100 000 but because everybody has do their parts well i will end this talk by by either not because it will go to 100 000 but because it will by itself thank you you