Ethereum Protocol Development: Past, Present and Future

Transcript

hi everyone my name is Tim Bako and I facilitate the awkward devs calls which are the calls where discussions and decisions about governance of the ethereum protocol happen and today I want to take some time to talk about my view of how ethereum governance works it's a subject for which there's not a lot of great resources and so hopefully going through it helps kind of paint a picture of how things are actually done on ethereum and as we start I think one important thing is to Define what do we talk about when we're discussing ethereum governance what's the actual saying that we're governing and I think there's basically two things

um that we we often refer to the first is what I've called like the ethereum protocol rules so effectively the ethereum network is based on a protocol which is a set of rules that are implemented in things like the ethereum yellow paper and eips or the python specifications that we have for the network that Define the sort of Step of functionalities and and things that a node has to do to participate on the ethereum network and so if you do not follow the rules of the protocols you're sort of kicked off of the network and that's really the main thing that you know when we say we're changing ethereum or we're we're sort of having an upgrade to the Ethereal Network this is the set

of things that are changing and and that affect that um and there are some other protocol rules that I've sort of categorized as nice to haves which are things that different clients will Implement and usually try to coordinate and agree on but it's not the end of the world if there's not like perfect match between all of this and these are things like the apis that the nodes expose like the beacon apis or Json rpcs uh some of the gossip protocols the sync protocols or the way that you know clients communicates with each other um so in practice you know we try to get everyone to agree to a standard here um but if there's not perfect alignment it doesn't mean that like a node will be

removed from the ethereum method um but so when we talk about you know governing you're changing the protocol rules these are effectively the two things that we need and more precisely I think when we talk about like ethereum governance people will refer to the ethereum mainnet um and so this is you know the the network like the uh ethereum network with chain id1 which has all of the balances that you know people considered or eat balances um and more importantly that has all of the assets the nfts The Tokens The Smart contracts and whatnot deployed on it um so when we talk about ethereum governance often we're concerned about you know what are the changes to the protocol that will be deployed to this instance of the ethereum network uh

rather than you know something like a test net or an L2 or an alternator one that might also Implement a lot of the protocol rules so at a high level um this is what you know what this is the thing that's being governed uh when we talk about ethereum governance the ethereum protocol rules as instantiated on the ethereum mainland and most of my talk now will be about who actually governs ethereum and how do the different stakeholders kind of relate to each other and I have this little diagram here which shows you sort of all these concentric circles that that kind of boil down to the ethereum protocol so you can think of the little image there as like you know the set of rules and the the state that we uh that we just

just referred to and there's different groups that each have uh that teach have an impact on ethereum governance and I think the closer you get to the protocol itself with things like core devs um maybe the more formal that role is but then as you scale out you get this more and more diffuse set of people who end up having kind of the majority of the weight that are able to kind of shift the entirety of the network one way or another um but in a less formalized and and kind of more diffuse way so we'll sort of start from the core here and make our way out and hopefully by the end of this you have a good view of like how governance works and how these different stakeholders kind of relate to each other so let's start with the chordifs um one

thing that's worth noting is you know there is no true definition of a core Dev this is often something that comes up um it's really like this informal mix of client developers researchers and other people who contribute to ethereum in practice we have over 10 client teams so these are you know formal teams that build uh software like the node software to connect to the network either spread out across a bunch of different types of organizations around the world uh there's nine of them their logos are on this slide and there's a couple more as well we have a bunch of researchers uh again from Academia from BC funds some andons some that work from the company the EF obviously has some as well and so on um and there's a last kind of group of

just random people who end up showing up on our Discord or somewhere else and we've tried more recently to like uh Force this process a bit more and bring like cohorts of new people every every now and then through something called the ethereum protocol Fellowship um but this really just like an open development culture which is allows people to come in and make contributions and I have four uh avatars of like random people who showed up on Discord over the years that are that are worth sort of zooming in on to to make this a bit more real um all the way to the left uh we have Danny Ryan which is someone who's who's quite well known so he came in as a contributor on some of the very early

proof of stake stuff totally sort of you know had no previous connections to ethereum and whatnot and today he effectively leads the research team at the EF and coordinates um all of the work around like proof of stake uh from from the EF side so he's like a great example of someone who came in in this open ecosystem uh made some extremely significant contributions over the years and now has a pretty major role um next up we have this like little eye Avatar this is POTUS uh someone who showed up as an Anon in the r d Discord with like a lot of strong opinions and and trying to like help improve the design of things and now he's joined uh the the Prismatic team uh so he was someone who just yeah through coming in

and contributing for a while like ended up being you know one of the core devs after that we have this like pineapple uh Avatar which is Proto Lambda so Prolo a bit similarly to Danny showed up to work on some of the research problems that were on ethereum um didn't know anyone just came and and you know ended up uh finding ended up finding a lot of solutions for problems that we had on the research side and um obviously has has helped with a lot of the scaling work uh So Pro he is the Proto for Proto bank sharding and now uh works on the optimism team uh helping the Scala theorem through l2s and then finally this last sabotage probably the least well known as someone called sea monkey so when we were when we were

doing all the testing leading up to the merge so ethereum's transitioned their proof of stake he was just some guy who showed up and was running every possible combination of clients so like you know there's nine here so like the uh the 20 pairwise combinations between the four and five he was running them all on his personal computer and just reporting bugs or ux issues as he saw them um and and sort of spent like the entire time of their merge doing this and another great example of just someone who came showed up and and contributed significantly to the protocol so formal definition of what a core Dev is I think my personal definition looks something like this where

um it's usually someone who's focused full time on ethereum protocol uh research or implementation and by full time I don't necessarily mean they're spending 40 hours a week but more that they're sticking over for a long period of time and and not only focus on a single EIP or feature right you don't just want to come in make a change and then go back uh go back to their other job they really have like a longer view over the protocol and this means you know those people often take responsibility for is ethereum secure is the theorem sound is it working as a whole well and over time these are the folks who like spend their time discussing and debating what changes should we make to the specs uh how should we Implement them and you know

should we potentially deploy this on on ethereum and I think it's really valuable to keep this uh as a like very loose definition and have this be a permissionless group um afri uh who who's like another core contribute it through ethereum recently posted this uh this message on a thread I believe around filecoin who wanted to make like a more closed and formal definition of quartets and I think I agree with pretty much everything he's highlighting there where um there's a lot of value in having this be be public and on the ethereum network a few things we've done for this is all of the calls that uh have to do with not only just like the core Dev calls but even calls about specific eips or or Side Tracks like for example you know

the 444 calls or eof calls as shown on this image they're all published at YouTube and live stream or at least made available with recordings so anyone can kind of follow up on them and then a lot of our discussion happens on public forums so whether this is uh ethereum magicians these research or we also have a Discord server and this is something that's also very important so that you know people can show up and get all of the context about something and and kind of understand what's going on so this is of course do in in this system and they're the ones who at the end of the day you know spend their time on a network develop a clients and propose changes to the ethereum number but they're not the only ones who can

propose a change so we have an open process for proposing changes to ethereum called the EIP process this stands for ethereum improvement proposal and there's a template and some instructions at eips.etherium.org um so anyone can go and you know just literally copy paste the EIP template and propose a change um and a lot of people just do this on their own especially for things that don't require protocol changes like application Level standards and whatnot um you see a lot of eips just like being driven by the community but I think you know for people who want to make a change that like significantly affects the ethereum network um you know it's a larger impactful change in in one way or another um and in practice those folks will end

up having to work with uh researchers or clients implementers or whatnot just to make sure that their their change actually works and is sound and you know is is kind of the best version it can be and so if you if you sort of signed up for this role um I think as an EIP Champion what you're trying to do you're trying to show that this is a valuable change that this is sound that you know and and you know won't break ethereum and as you're doing that sort of creating momentum around the change to you know signal that this is something that should happen and that's important and then practice what happens is you come up with this idea to change ethereum um you you bring it to researchers or client developers and they'll find an issue with it they'll find usually a Dos

Vector you know something that could like halt or stall on ethereum and then the the best kind of uh the best kind of eips will spend time to address all of these and iterate on their proposal to get to a spot where um it is safe and you know this is a second screenshot is just a search for dos I'm a theater magicians and you can see uh literally just from the top results uh account abstraction VIPs uh protobank Chardon and then the withdrawals VIPs all have you know threads talking about how to deal with the Dos vectors that they would introduce um but then if you do a really good job you end up at a spot where like okay you've done all these things and one recent uh example I think did a really

good job was uh the folks from uniswap who checked in EIP 1153 where you know they implemented it across all of the clients wrote extensive testing um got some like feedback from not only core developers language developers and also from the community to kind of show that there was value for this um and this was something where even though there wasn't like uh there were still some objections and not everybody you know is is necessarily uh it's not necessarily everybody's favorite proposal um I think the quality of the work that they did the feedback that they got and then you know the support that they they managed to show from the community uh meant that it got scheduled for the next Network upgrade um so and this is something that took

them you know uh well over a year I think um since the time that they started uh approaching this to the time that like it's actually gonna get out of there but so if you have you know these people that like write the specs and codes and and write down the clients the question that always comes up is like don't EIP Champions and core devs just control ethereum and my answer to this is that no they do not if we if we look at again our our chart that I showed earlier um I think you often get this perception mostly because EIP Champions and cordovs are the ones that are the closest to the actual core protocol and who have the

most formalized role in the process in a way um and even maybe another way to put this is that not only is their role formalized but it's like advertised as such um and so it's it's very easy to like end up with this perception that they're the ones who sort of get to choose everything um but the rest of this talk will walk you through what everybody else does and how they play a role um and why this is not the case so first off looking at like consensus participants and these are you know usually either minors or validators so on ethereum right now these are the proof of stake validators if you have an upgrade so you know say the core devs and the EIP Champions agree to like a set of changes to go live on ethereum it's necessary for at least some of the

consensus participants to run the upgrade otherwise you know it never happens like it does the chain doesn't upgrade automatically you need the people to manually go and um change their version of the software to do this um but this is not sufficient alone right because if only a subset of their consensus participants upgrade so if you don't get anyone what happens on ethereum is the chain splits into two and so there'll be people who like run the version of the upgrades people who don't and under the proof of stake rules the validators on each side will just slowly uh burn the stake of the validators who have not upgraded on their view of the chain until those validators represent less than a third of the stake and then the chain will keep finding realizing again so both sides would sort of you know gradually

uh destroy the ether of the validators who aren't on their view of the chain so that they can continue and keep going um but if if nobody upgrades it yeah you don't even have this you know potential risk of a fork in the first place because the upgrade has never gone live and just to put this in practice uh or numbers sorry as well uh if you look at like core devs as an example um cordovs don't run a significant amount of consensus nodes on the network so a few years ago we announced uh a client incentive program where we gave all of the client teams uh about 144 validators to uh to to run and obviously test their software to make sure that when we're telling everybody else around

those elevators we know it's working so imagine you know we have 10 teams they get a uh 144 validators that's roughly 1.4 000 validators um but assume you know that they were they had some secret validators right like that they're they're not part of the the program that was announced by the EF say they had 10x more than that so you go from like 1.4k to like 14k validators currently there's about 560 000 validators on the ethereum network so that would represent only two and a half percent of the entire number so you can imagine a world where like the core devs you know single-handedly try to push something well it means that like 97.5 percent of the network in this like extreme scenario where they have 10x

more hidden validators in what is public um still they don't even get like three percent of the network so you know the vast majority of the stake is not controlled by them and you could imagine you know what if the EF sort of joined forces with them so uh we coordinate a lot of these changes so maybe we we decide to do this um so the ef's balance is public you know we have about 300 000 each um this is not staked so we would have to activate all of this as validators it should be public Unchained or it would you know cause a delay and whatnot but imagine we get there um then we combine with all the client teams we get about 24K validators which is like 4.4 of the network today which is still

like a very very minor share of the network and to put this in perspective we usually have somewhere between 97 and 99 of validators online depending even time so it's like this scenario where the EF goes all in uh the Val the clients have some secret stash of eats that they're creating validators with um gets you roughly like 50 more than what can be offline on a given day so it's still like an extremely minor share of the network um and this this is just you know the emphasize that like VIP Champions courtevs the ethereum foundation can all get together and you know propose a change run it on all their nodes but it still needs the overwhelming majority of the rest of the state to to uh to

support this change if it is to be the majority Fork um but then even if that happens um consensus participants so like the validators aren't actually the ones who get a final say on ethereum and this is where we're we're expanding out from the circle again so next up um node operators are extremely important on ethereum and this is why we spent so much time talking about decentralization and why it's important to be able to run a node in all of this and the reason is that on ethereum every single node validates every single transaction and as an aside here this like validator is a terrible term for what the consensus participants on ethereum do because of this it's sort of offuscates that every node actually

validates transactions so validators are the only one who do this and and if you imagine a world where only the validators Run nodes then they could collude and change their protocol rules without being noticed by anyone right imagine a blockchain where it's like me and one person running the validator nodes and we're the only one who run notes nobody else can then I can mint myself you know a million tokens so can my friend and nobody else is going to know that the supply chains because they can't actually scan the chain or run a note to know and this is why it's important where the more people run those the more people independently verify that the rules of the protocol was were followed and you know the example I gave with me and my friend is an example that's like based on malice like we collude we do

something but you could also Imagine like an example where me and my friend are running the nodes but we do everything right but we have a bug because they're both in my apartment and you know something bad happens or we run a buggy version of the software so a small validator set also increases the risks of coordinated bugs and this is something that the non-validator nodes would would notice as well um and so if you have a a blockchain where you know the non-validator nodes are abundant then the game theory for the validators is if they propose a block that is invalid because you know either they meant themselves infinite eth or they uh they have a bug in it then all of the network then the other nodes sorry on the network will reject

it when they see it and it just won't propagate it and this means it's validator will have like missed their slot right they'll just have proposed this block that was bad everybody will agree that it's bad and they'll miss the potential fees they could have gotten on a valid block and so so it's like very important to run a note because when you do so you're effectively sharing your view of what is correct with all of the other nodes that you pair with on the ethereum network this effective activities USA in governance because it allows you to you know protect against uh the consensus participants trying to like feed blocks that change their protocol rules they don't get they don't get to like create new protocol rules and have all of the other nodes agree to them they can only you know run the software that they want

but then they need the entire network to accept those blocks but even Beyond you know writing a note on the network we can think about like the users of ethereum and one version of this question that comes up uh every now and then is you know can a single project decide which ethereum four quins imagine you know there's a contentious Fork um is usdc here is maker or is anyone else going to be like determinant in like which is the real Fork it's sort of explained kind of how a fork would work on ethereum here um so if we have two potential uh two potential Forks so imagine there's a proposed upgrade some of the community upgrades to it because they want it's

the other parts does not then at that point of the upgrade the entire state of the network is duplicated across both sides um so every account balance every contract every token nft and so on there'll be a copy like an identical copy at the fork on both sides of the chain and for some things this is fine right like so you know the native assets of the chain like eth for example can live on both forks and they might have just a fluctuating value and you've seen this in the past with like Ethan eth classics there's also some Unchained nfts that you know might be totally fine if all the metadata is stored on shame whether the the network Forks you know the metadata is stored on both sides that's totally fine but then there's some there's some assets to some projects that actually need to make a choice

because they're referring to something that's not like fully encapsulated within ethereum and therefore they need to choose where is this thing where is this pointer going to point to so think of something like real world assets including stable coins right there's one dollar in a bank if there's two forks you know they need to make a decision of like which of those Forks can redeem for a dollar or similarly if there's an oracle that's saying like you know what the price of eth is well they need to figure out you know like the price of East on coinbase and binance that gets mapped back on chain to ethereum what chain is that to what's you know off-chain pick your asset and similar thing for Bridges you know imagine like an L2 Bridge or a bridge or

alt layer one if they're taking eth from ethereum or if they have Bridge tokens you know imagine like a bridge L2 token on East they need to recognize okay which of those Forks is the one that will uh honor a bridged asset from and so um as a whole all you know the entire ecosystem like all these projects coordinating together on a fork can like ensure that all of the assets on it migrate with minimal issue and and sort of increase the odds that that is the fork that's recognized as like ethereum but you can't have like a single project this right like if only usdc or only the arbitrary Bridge are only crypto punks you know where to say this random Fork is the ethereum fork everybody else

would probably just say no um but this is where you know the this is where the the the the the the power becomes much stronger but much more diffused where if you have this like shelling point on a single chain then it's extremely powerful and and it helps kind of tilt uh tilt the fork towards this chain being the real ethereum and you can imagine a world where like you know the validators go Rogue and there's not enough people running nodes and you still have you know all of the projects recognizing your single fork and this brings me to my final Point uh which is you know quote-unquote the community or I believe that bankless people call this layer zero so at the end of the day ethereum

and you know a bunch of other things like Bitcoin like the dollar are all just a shared reality right they're this thing that we agree to as having all of these properties um but that can change and we have processes you know that are in place on ethereum to make changes in like a happy case so like in a in like a more formalized everything goes well type of way this is what we we sort of touched on you know the EIP process Network upgrades and whatnot um but there are there are also ways to change a network in more extreme scenarios and these are things like contentious sports like potentially social slashing and whatnot where the community has to like come together and agree to what is the right outcome right like if we have a contentious Fork which one actually is ethereum if there is

like a social slashing you know how do we go about this and as you have these these uh these more extreme cases then the second order impacts become really important right because it becomes a question of like what are the values that we want ethereum to have you know if we want to have credible neutrality then you know is there one of the forks that like affects this to an extent versus not um if we want to have like platform stability like if we want ethereum to be somewhere where we expect applications to you know stay and and be used for a long time like you know is is like a contentious fork or a social slashing coming at a cost to that um and there's no there's no Pro caller there's no like clean way to resolve those things but they are at the end of

the day they're more powerful right like the community is like the most powerful most diffuse things where we can all come to consensus on what is ethereum and you know what the trade-off space is in a given situation uh where you know there is like an extreme case like that's just work or something like that and this is possible because these systems are permissionless and open right like if all of ethereum was closed source and you know it wasn't possible to like uh actually take the code and run with that and make the change or whatnot and you couldn't do it and one example I think that's interesting if that a bit beyond ethereum is like Bitcoin right there was this big Bitcoin cash fork and even today the two websites bitcoin.com and bitcoin.org sort of resolved to two different uh two

different projects so Bitcoin cash and and you know Bitcoin as it's uh mostly known um and you know whatever you think of Bitcoin cash versus Bitcoin like the fact that this is possible that like you can have literally you know two top level domains like uh or you know two like Prime domains redirecting to like different versions of Bitcoin kind of shows you like how powerful this permissionlessness aspect is of the openness of these systems are um and and and you know like the sort of resulting effects when like part of the community decides to take over and go in a computer different direction so you know hopefully this gives you an idea of like who are all the different

stakeholders in this and um how they come together and and the impact that they can have on the process now I I just want to talk a bit more about like some specific cases and uh you know just some high level thoughts about like the future of ethereum specifically so one question that that comes up a lot is like why even if all of this is true like why don't we feel it right like why does it feel like potentially you know court has a VIP uh Champions get to like make these decisions and like the rest of the community doesn't quite get a say and my answer to this would be that if core devs are doing their job well then they'll listen to the community and they won't push forward a contentious upgrade

and they have an incentive to do this because if they push a contentious upgrade and have to deal with you know all of the messiness that comes with a contentious work they're the ones who actually have to like write the software for all of this um and so the cost of like dealing with a contentious Fork is very much borne by by these people and that I think helps like make them hyper aware of you know not pushing anything that might be to this and if you look in practice like in ethereum's history um I think there were four probably big contentious cases and you know I'd say probably more three into fourth is is more of like a a an extra example um but definitely the most contentious one was the Dao so this was uh very early in ethereum's history about like a

year after ethereum has gone live uh there was a hack which had uh 15 of the East kind of affected by it and there there were a lot of uh differing opinions about how should we approach this you know should we recover the funds should we not and in this case what cordovs did is they put out uh software which allowed users to Signal whether or not they wanted to run this upgrade which would return people's money um and so this was you know I think a pretty neutral approach um and then most of the most of the community sort of rallied around uh on reversing the hack and uh a small part of the community didn't though and this is where ethereum classic comes from

right the sort of the version of this of ethereum or this hack happened and you know was not uh was not changed after the fact but then a couple years after there's another big hack uh or a bug actually um on the second most used uh on a multi-sig which was uh written by the parity team which had the second most used client after gath and was also led by ethereum's co-founder Gavin um so it literally made all of their fun stuff as well as I believe it's like a hundred letter projects funds were stuck and there were a couple VIPs to resolve this most notably eip999 and this got a ton of community opposition and so even

though this was proposed by literally like one of the co-founders of ethereum um it was one of the you know the second most used cordov uh the skin most used clients so like you know a very Central core Dev team um the rest of the the different kind teams uh never sort of agreed to propose this in an upgrade due to the community opposition so they're this was even like a more conservative stance than was taken with the Dao and then after that another VIP that came up with EIP 1057 uh also known as Prague pal so this was a change to how ethereum's proof of work algorithm Works where um ethereum has always had this goal as stated in its yellow paper to make the network Asic resistant uh so to

not end up with like uh basically private uh companies building hardware for it and be more accessible to gpus we saw this starting to change so people propose an upgrade to the proof of work mechanism to try and uh re-tilt things towards gpus and there was a lot of community pushback against this as well and I think this is something where um it's hard to say exactly but quarters were probably like weekly in favor of it um but seeing the strong Community push back let them to not put this in an upgrade and so Prague pal also wasn't ever shipped and then my last example is a bit different but on the merge right like ethereum's transition from proof of work to proven stake this was part of ethereum's you know quote-unquote roadmaps since the very early days

um in as much as there is a roadmap for ethereum um there was obviously pretty strong community support and it's worth noting that like the ethereum classic Fork that resulted a few years ago I think helped move a lot of like the very proof of work people uh to a different Community but miners were obviously very opposed to this like we were taking away their entire business um so this is an example where you know the client teams clearly push forward and propose the upgrade um I'd say the overwhelming majority of the the community transitioned but there is like this tiny minority Fork that's popped up uh around the time of the merge ethereum W and I believe it still exists today although it's not like maintained much um but it is you know an example of like

part of the community was able to opt out of this and keep doing their own thing even though the vast majority kind of moved in One Direction and I I think that assessment is what led the core devs you know feeling comfortable pushing the software knowing that like you know there'd be very minimal uh parts of the community who thought this was contentious so what all this says you know know it's worth maybe giving a high level recap of like what's good and what's bad about ethereum's governance structure one thing you know that's great is like permissionlessness everyone can participate in it uh you don't need to ask anyone you just you can just sort of show up and and contribute but one downside of these Stripes structures is uh you know best encapsulated in this as

a the tyranny of structurelessness um so it's it's much harder to like make changes and and um affect actual change in a structure like this so if you haven't read this essay I would recommend it as a good view of like the failure modes of something like uh this permission this governance another Pro is that it is a capture resistant process so because the process is very informal and and malleable it means that if it tries to get captured or attacked um we can sort of change the rules of the process quite easily where if there was you know like a very uh formal on-chain structure that you know you have to do X and Y and Z to get an upgrade and you're sort of telling people if you do X and Y and Z you can capture the protocol

um so because we have a Vigor process I think this helps uh be more responsive but this comes with the downside that like if you don't know about the process it's extremely hard to navigate and you need to build up all of this context um to be effective within it and then lastly uh you know there are several checks and balances so as we went through this entire kind of bunion of different stakeholders and these are all effectively checks on one another which is which is very valuable because it means one of those groups can't like unilaterally change things in One Direction but this means that when there are contentious issues it's really hard to gauge what the community thinks right like what's the what's the consensus here how do we weight different stakeholders there's no there's no formal way or like unique way to do that

and so um yeah I'll wrap up with some of my thoughts on like you know where we're going from here um and and sort of what I've seen over the past couple years one thing I think is interesting is there's more and more people who like have a seat or want to see it at the table um there's a lot of teams that are not only like affected by changes to ethereum because they've built businesses or protocols on it but that also want to help shape things you know so like l2s uh infrastructure major adopts we mentioned unisoft earlier they sort of want to contribute so um there's many more people who like get involved in this and then in parallel to that a lot of the easy stuff has been done so I have this High hanging fruits uh I guess phrase where most of the low hanging

fruits have been picked and a lot of our new proposals like have trade-offs and you can have you know reasonable positions on both sides of those trade-offs so it's it's much harder to like say oh this is the obviously right thing and this is the obviously wrong thing and so if you combine those two things right like if there's more people you need to align and like more difficult trade-offs I think it does make it harder to move forward right like it slows things down by default and add to that that there is more at risk because there's just more value like built on ethereum there's more applications you know more things we don't want to break um so one big open question that you know I have is like does this lead to ossification by default does this mean that over time it just gets harder and

harder to change theorem and you know at some point it's like so hard that like you literally can't um and I don't have an answer here but it's something we're thinking about and a few things um that you know myself and others have been uh trying to do to help um make things I'd say a bit better on the governance uh point of view in the in the past uh year or so one is just aligning uh how the specifications are done on the consensus and execution layer so since the merge um you know those things technically are are the same uh or you know talk to each other uh but they have like very different uh processes so trying to get that to like a single process to make it

easier for people to understand changes that are being proposed to ethereum and also to propose a change themselves if they want so that's uh work that's been in progress uh the other thing is I think more and more trying to use like async forums like East magicians um to uh gather feedback to like signal changes um that not only like are a bit more open to people coming in whenever um but also like less ephemeral than things like using the calls or agendas for the calls which which are extremely hard to then uh to then dig into and then this is a bit more of a wild bet but and you have the Summers funding this initiative called Summer referral calls which uh basically is funding 30 researchers from a bunch of different

domains from Healthcare to like environments to like seven blockchain and Tech stuff to think about protocols generally some of the hard questions about them and see if we can come up with better mental models better Frameworks to address these and then lastly uh I'll leave you with some stuff I don't have an answer for and that's kind of on my to-do list um so balancing you know speed versus stability of the network kind of related to my previous slide but you know if we know we're gonna ossify we should probably go quicker but if there's more at stake we should probably be more careful with how we approach things um so finding the right balance between those two things I think similarly um we have this big focus on l2s on ethereum this is the way we see ourselves scaling the protocol so um

what should it look like for l2s to standardize should they try to like you know come to L1 or through L1 governance for some of that process if so which parts uh what does that look like and is there a trade-off like you know there's too much standardization then it's kind of used as to have many l2s you want like a diversity of approaches but are there some things that are worth you know unifying or potentially even enshrining in L1 and then my last uh big open question is like uh ethereum governance works you know mostly uh like the formal process works mostly through like these calls that we have um you know can we do even more to to move things async and have less of a focus on that um the good thing about calls is they

give us a lot of momentum but the bad thing is you know they optimize for people who can like show up on the call speak in English think on their feet and and whatnot which is not necessarily the set of the things we want to optimize for uh for long-term governance and so how do we like slowly transition that over come on and so yeah that's all I had uh thank you so much for listening hopefully uh this was helpful and kind of helped you understand a bit how ethereum governance Works uh if you have any questions you can reach out to me on Twitter at Tim bako or feel free to email me I'm Tim ethereum.org and yeah it was nice uh presenting this to all of you cheers