What's new in Eth2 - 01 July 2022 with Ben Edgington and superphiz

Transcript

hey everyone i'm superfiz uh and i'm here with my great friend ben edgington every fortnight we record a a kind of a walkthrough of ben's newsletter what's new and eve too hey ben hey fizz good to see you again happy uh independence day though it somewhat pains me to to wish you this oh no hey you know what it's uh thank you um so ben and i have been doing this for uh what probably 10 or 12 weeks now something like that or 10 or 12 fortnights do you call them is that right is that how i'd say it i don't know i don't know um yeah but it's been a few months it's been a few

months okay um and one of the things i i kind of noticed is is we don't we don't do much of an introduction and i know most of you don't care about an introduction you just want to get into the newsletter and so i think maybe uh we will reach for a happy medium where i kind of say hey i'm super fizz i'm a community member um i work with an organization called eath staker and our goal is to um sort of promote uh staking in any way that people can stake with an emphasis on solo staking um you can find me if you ever want to hang out um i spend a lot of time in reddit.com r e finance um i spent a lot

of time in uh the rocket pool trading discord uh and lately i have developed a a terrible twitter habit uh so i'm at superfiz at twitter so if you'd ever like to connect that's where you can find me ben how about you so i am uh for a day job um product lead for the tecu staking client so the ethereum staking client i am also a solo staker and uh consider the e-staker community to be kind of family um and i write this what's unique to thing every couple of weeks been doing that for about four years now which is what we're here to talk about awesome let's

i was gonna say let's get into it do you know who says that um that is anthony cesano's uh trigger for his his episode now let me figure out how to share this screen a little behind here there we go all right so we're doing what's new in e2 for the first of july and today is actually the fourth of july 2022. um i guess you publish these on thursdays and we go over them on mondays which gives me three days to think about uh reading it and annotating it before the 15 minutes before the call i actually do it yeah well i'm sure we're both great at

procrastinating i mean it's actually friday night i get this out tonight yeah i plan to get ahead of it yeah all week i'm thinking okay let's make some program make some progress and it gets to about 8 pm on a friday and i think okay better get started do you make mental notes or physical notes like i'm sure you have uh i have a trello board so uh the interesting thing about trello is it's really easy to send stuff to so if i come across an article webpage or twitter or reddit or whatever it's very easy just to bang it into trello one way or another and then at the end of the week or you know when it's time to start compiling this thing i just go through the boards and move things across columns and uh process it all and try and weave some sort of

narrative of the uh of the week out of it all yeah that's really a great tidbit like um i have i have tinkered with trello but i really do see how it could be a valuable resource here are there plugins for it or do you actually have to go to trello because i said it was so easy uh usually so uh i can share stuff directly from my android phone i can't speak for other types of phone but you know i can on most pages as a share button and trello is one of the options that pops up automatically i'm you know a bit of a techno favorite thing about this stuff but it just works you know i can make a new card and add it to the list i'm just happy you're an android user uh i that's

i i have uh okay i don't tell anyone this this is a secret between you and i but i i uh look down my nose at people who use iphones so but that's that's kind of a secret so come possibly coming so yeah i was on layer zero so you listened i i urged you not to it was good uh that was really good fizz i um very much enjoyed that and but just your clarity of thought around decentralization but also um the drive i mean really thoughtful answers to questions like we had this whole lido thing about should they self-limit and it was like well if they self-limit won't others

come up and you know just take their crown and whatever and and and your response thought was really good you know we will deal with those if and when it happens but you know everyone should should take responsibility for doing the right thing themselves not look at others um yeah and yeah i just thought that was great so i i haven't actually said this publicly but my my thoughts on lido have have really evolved um they haven't changed but um when they have have made such a firm and strong position that that they have no interest in self-limiting um i i believe then it becomes the responsibility of the protocol to act um and if if layer zero if the community isn't a lever that can change that behavior um

then it's time for uh us to step back and say it's it's time for you know the protocol to act and i i that gives me a lot of peace at night knowing that um you know it's i'm not responsible for this the community isn't responsible for everything we're responsible for doing our part so uh yeah that's just a fun thing to share uh yeah i'm a bit um intrigued by what you mean by the protocol um in this respect we we need to build in mechanisms in into the clients that that prevent this kind of centralization is that where you're yeah i'm not you know i'm not a uh i don't have any vindictiveness in me i'm not a

vindictive person but i do believe that the protocol needs needs to find ways to limit the kind of centralization that that is perceived as an attack on the protocol um yeah i mean i think the interesting thing with lido is that would be extremely difficult because they are represented by 29 different operators so you know add any kind of protocol level non-governance level they look like completely distributed entities so i think in in protocol that would be impossible without some metadata you know saying that actually this is a an organization that's comprised of many individual organizations without that

sort of metadata that would be really hard sure and so one of the things that i want to do is let me think of how to phrase it like i i have ideas but i don't think it's i'm not interested in pursuing anything that's not just a community groundswell effort and so when it gets to protocol development my my place is to trust protocol developers to solve things that can only be solved at the protocol level and if it's something that can be solved at the social layer then

it's something that you know i'm i'm all about um but in the same way like you kind of look to me and you're like hey fizz you represent the community and i'm like hey i look to protocol developers and i'm like hey you represent the protocol so i'm trying to walk this fine line of of not making it my place to say or even the community's place to say this is what ought to happen um it is we are ethereans we uh subscribe to the ethereum network and we believe that there are healthy solutions that promote decentralization um that's i guess kind of where i'm at cool cool yep yep let's go

it's funny i was like we're just going to gloss over this one um yeah so matt cutler ethereum ethereum's hidden power structures um another thing i'm very interested in what'd you get from that yeah um it's a very good overview this is another podcast um very good overview of what the post-merge block production landscape is going to look like so you may be aware i'm sure you're aware is that there's this because of mev right it all it all comes back to mev and the um dominance that has within in the protocol uh and there's there's a

problem if solo stakers like yourself and myself can't access mev it you know that's a big pressure for us to throw a lot in with a big staking provider um because they they will have access to this this stuff um and uh we'll get better returns so there's definitely a drive to try to um democratize the mev and the mvv is going to happen right i mean you know we might take individual stands against it or principled stands and say we're not going to participate we're just going to build our own blocks but reality is for the majority of people they're going to want that potential revenue so uh the flashbots have been designing what they call mev boost mev boost which is um a

whole architecture for block builders to communicate with validators via a relay and the relay is trusted so this is a sort of flaw in the design in a sense and in future we might be able to do this in protocol and remove the trust from from the equation but for the time being you have this sort of somewhat trusted relay um and solo validators can uh ask a relay for blocks when it's their turn to to propose and behind that there's this block builder ecosystem and um yeah what was really interesting about [Music] matt cutler's episode was just exploring the different roles you've got the searches you find mev you've got different kinds of block

builders who will be motivated by different things there's kind of potentially a whole ecosystem of block builders what i've been really worried about is that uh effectively there'll only be one block builder on the network flashbox will be the the sole supplier of blocks to the network and that's bad from a single point failure point of view and it's bad from a sort of centralized censorship resistance point of view as well so the the best solution for this is to have an ecosystem of block builders um all competing against each other i i'm somewhat more encouraged after a couple of meetings i've had and listening to some podcasts and things

like this one that that ecosystem of block builders might actually become viable might spring up and provide a healthy diversity so yeah that's good that's great so i i have kind of said for a few months now that i believe the three biggest threats to the future of ethereum are um lsd or liquid staking derivative centralization um they are mev centralized clients where one client fails and disrupts everything and so i i have publicly addressed ls decentralization and i've publicly addressed client centralization but mev

just scares me um and i i haven't it's funny because it's it's in many ways a necessary evil um well you know i i can say i am fizz and i am not going to participate in mev and it doesn't change anything it's it's still gonna go on so there's no point in taking this you know noble stand we really have to find a way to um deal with it as a community yeah uh when we have this in protocol and there are proposals and and ways that this might be done and make it trustless then there are interesting things we might look at like mev smoothing where instead of you know you get one block that's worth suddenly a hundred each because of some huge

arbitrage and then um and it's an outlier um for one lucky validator you know you can potentially spread that reward across many validators and just make the whole thing much more um a fair as it were how many of us as validators aren't just like secretly hoping to hit that 100 mev lottery uh i mean it's true like i i'm still like maybe i'll hit one uh and it's funny like that's the point where i acknowledge we are all greedy so and it's not unhealthy greed it's just a thing that needs to be managed yeah my missus in uh the face of the impending merge has been uh cranking up the number of validators on her staking rig um

taking advantage of uh current eat prices sharp lady [Laughter] exactly mainnet shadow fork seven this is a really interesting one yeah um kind of i mean the shadow forks have sort of become almost routine now i mean when when we started these things they seemed huge uh like massive events and we had that um whole day when we were in amsterdam where we got together as all the dev teams i think it was number three um and uh sort of nurse made the shadow fork and there were film crews there two two separate film crews making documentaries and all of this and now it's like yeah okay it's happening tomorrow fine whatever i'll look at the results later you know it's um uh

they're cranking through and going pretty well i mean there are niggles right and there there are always going to be niggles it's never ever going to go perfectly the question is are these just niggles or are they fundamental flaws are they unfixable or are they a lot of work to fix are they indicating that some client is really you know behind the curve in terms of his readiness and the answer is no i mean there are little issues here and there which are configuration issues or unrelated to merge issues um and uh and showing up you know just maybe minor easily fixed flaws in the in the spec we we had to change a a may to assured in the spec you know because um the the may was too loose and

uh uh some clients were were not retrying some calls when perhaps they should have been doing so so you know it's true trivia like that so so we're getting there i mean we'd still like to see completely clean shadow forks um happening but uh i'm pretty content with the way they're going yeah i you know i see issues here with air gonna basu um and one of the and also with nethermine i it's interesting because i'm very eager to promote execution diversity execution labor diverse execution client diversity um but i just don't have that that confidence that i have had for the consensus

clients that that i can look someone in the eye and say run run this secondary client and and it will be fine like no worries i just haven't developed that sense of confidence and i really do hope that that comes in the next few weeks yeah yeah i think there is a slight maturity issue um you know it would be perfect if everything was completely production ready and confidence inspiring before the merge but uh it's not a it's not a cliff right you know we can update clients later on as as they catch up um basically is chasing down this um they've got this new bonsai tree database storage format which is brilliant i mean it's it's you know

beats guess format you don't have to prune it manually it's auto pruning uh and all of that but there's some really subtle concurrency issue where you know threading issue where the database gets corrupted totally unrelated to the merge the merge somehow seems to change the timings to make this show up more frequently which is why a lot of their notes fell over during the shadow fork seven um they are you know all hands on deck um they've got a half a dozen people like drilling down on this currently um to to fix once that's fixed i'm going to be um pretty confident that basu will be good to go but uh yeah until until the next thing i think you know for my part it is important for me to encourage people to

use uh basic eric on another mind um i haven't i haven't you know shouted it from the mountaintops yet but we do need people who maybe have more opportunity to attend to their nodes or maybe more technical expertise to run those nodes because there is a potential for guest default and we we will need um other nodes to to carry the network in the unlikely event that does happen so um yeah i've just seen a typo this is terrible i've played i blame them take you never mind issue on eregon i better update that i'll scroll fast um so another shadow fork tomorrow i

wasn't aware of shadow fork eight but i've been out of the loop so yeah as i say i mean um parry just puts up the configs and uh and and gets on with where client teams are are barely involved um so it's um it's happening and we'll we'll see the results cool here's a big one for each staker uh well for everyone i guess oh yeah uh sapolia is going to merge uh and it's funny if you look at bordeaux let's peek at it real quick they have targeted this so well at the moment like look at that it is they've been uh tinkering with the hash rate and you can see tinkering and they are the goal is to line up the red and the green dot to to get this uh right at

the the plan ttd july 6th yeah when i wrote on friday it was looking like the fifth but uh the target is actually the sixth so uh that looks pretty good yeah i kind of uh i guess here seeing these i i i was like you know posting we're gonna merge in 10 minutes but didn't happen thankfully oh boy i'll be so glad to get rid of proof of work that's another unknown yes this random timing stuff uh and so so polio uh e-sticker's doing a live stream with anthony sisano that's going to be on the east acre youtube channel it'll be similar to the ropsten call except this time it will be led by anthony uh and so i'm

really looking forward to that yeah i hope i can join you for that yeah i always love to have you there what's the favorite footnote that's a cheeky one ah yes this is did you watch the expanse no no you should watch the expanse this is uh this is great um sci-fi kind of um soap opera stuff it's uh um it's great uh very uh very good over the last few years but uh yeah it's uh this is just a scene from the expanse but we we didn't go through it but basically he uh he decelerates very violently i think it's right wow okay

and you know this just reminds me the last three four years all the you know consensus devs the uh we've all been uh working at a furious rate basically without any serious regard to governance we've been you know very lightweight in terms of the governance we've just kind of done our own thing and now we're having to deal with the whole kind of eat one governor's side and we're just not used to how slowly things happen an um like amusement community amusement for like you know six years and now it's i can oh i get the whole going you know like the speed of light and then slowing to the pace of a snail just

hitting a wall um but you know i mean it's uh i think tim's doing good work over there just um yeah making keeping things moving and uh you know i don't think it will be too serious it's just a funny observation that we'd agreed one thing on the consensus call and on the all-called devs they completely ignore that and agree something totally different but yep rapston sync tests um let's see yeah this is just happening in the background so um syncing turns out that basically you've got two independent clients with two separate blockchains you've got post merge you've still got the

um the eth1 chain if i may call it out i know i should stop saying that you've got the ether1 chain and you've got the consensus speaking chain say execution chain oh yeah that'll do all right the execution chain you've got the consensus chain and and they are tightly coupled together but um you've got to sync them up separately uh and this is basically because uh validating the execution blocks is a slow part of the operation so we have this optimistic sync whereby the beacon chain kind of syncs itself up and ignores the execution blocks um and then it goes to the execution client and says okay we think this is the chain of blocks that you need to validate and so validate it because at that point

the consensus client doesn't know if all the blocks are valid in the chain that it it has so it's kind of complicated because you you you've effectively got two different sync operations thinking on two different networks uh happening uh there so yeah it's good that um this is being given uh a full um run around on the uh uh so the ef team are doing that yeah i kind of got sidetracked trying to process how that would work in my head yeah you know one of the things i end up doing is um i end up taking some design elements for granted like um saying you know what they're they're

gonna figure that out i don't have to know that and then i recognize there are some things that i shouldn't have that i should understand that maybe i don't put enough attention into and i think that that dual sync is is one thing that i would like to understand better indeed well me too so technical bits engine api sequence diagrams yeah nothing significant um somebody was asking for this on twitter so i just thought i'd put the latest in um just this is how the consensus uh client and the execution client speak to each other um i found that pretty this is pretty

interesting geth running without a consensus client yeah um so this is the idea that the consensus layer you can the beacon chain you can basically just take um its lead you can basically trust that the consensus chain is doing its right thing and just take a very small amount of information from it and keep your execution side in in sync so the i'm not exactly sure what peter is is proposing here this seems to fall short of attaching a like client to it the there's the right way to do this the kind of the purest way to do this is to have a light client locally which

validates um the beacon chain or it proves um uh the state of the beacon chain and then that can drive your local execution clients so it's a it's a very nice um potentially light lightweight to to do that if you just want to run adapt you don't want to validate the chain you know you don't want to be a validator do staking on your on your client you just want the rp um the json rpc api and and so on so it's potentially a good way forward if you you know people who are just running e1 clients now can just add a sidecar sort of like client and then

and carry on as they are without putting a whole beacon node on there nimbus team as a good thread today from yatek from nimbus team on light client stuff and what they're doing at nimbus to build uh like clients so it's happening it's coming it's um this is definitely going to be feasible it's interesting because when i first began to conceptualize the merge i imagined or at least the current merge product um i i imagined that uh there would be kind of rolled up clients like the tecu basic client and and you might install that as as one system and this seems like a divergence from that and i think the reality is probably going to be in the middle where we'll continue to use two separately for quite some time

yeah i mean uh just to sort of give a little insight we are talking internally about you know a a tecu basic client and and what that might look like and how we might get there you know we're one of the few organizations that go has got both the production you know ethon and e3 client in in-house so it's natural and i it depends what you're doing and it depends on what kind of trust model that you're prepared to tolerate so um yeah the gold standard will be to to run a full consensus client and a full execution client and validate everything but there will be degrees between that you know from in-browser like client stuff where you um have the trust assumptions of a like client which are weaker

rather kind of stronger you know the way you talk about trust models is kind of backwards they're kind of stronger which means worse you know you you trust it less uh than um than a than a full node and so forth so uh there will be a spectrum and an array and there will no doubt be endless twitter debates and lots of sniping from bitcoiners about about this uh uh in the years to come but uh there's going to be a an ecosystem of different uh uh ways we can we can structure this you know my favorite part about this is as so let's let's imagine that tecumbe bundle up and then you know many people run that bundle there will always be an incentive to run

the the least represented of those bundles whether that be something like nimbus and aragon because you your individual penalties for being offline will be so small but you still have that potential for one of the other major clients to fault and then you're the the one who gets to to maintain the network so i think there's always going to be that uh that incentive to run a an underrepresented client yeah yeah i hope so and yeah to be clear we're not talking at all about um completely merging then we'll we'll always maintain taking base with separate clients that compare up with other counterparts it's fair to say

bundling them is that is that okay yeah it's an ongoing conversation i mean it's we will see i think it will be a process of just moving closer together in terms of you know things like unifying dependencies unifying documentation unifying metrics unifying logging and just generally moving them closer together yeah packaging together and then kind of emerging code bases a little bit keep them separate some of that's just natural over time like i think it's a natural progression what you know no matter what uh staking did you catch this call i was uh on some of it i got to it late um but uh is i'm continually impressed by the the quality

of uh of um what's going on in in in this world um you know this is done well uh call number 17 is is impressive these things sometimes start and they fizzle out after two three calls you know three four calls because nobody's really committed or interested but uh steakhouse has really got momentum and um there's a terrific core set of contributors who are building tooling for you know individual stakers to enable them to get up and running um as frictionlessly as possible uh and it's it's terrific um there's some really good work going on here yeah and that's i i love that you point out that consistency because we have we've tried a lot of initiatives and

many of them just fall short uh but you know colfax picked up steakhouse and ran with it and then remy uh remy roy picked it up after that and these guys are just metronomes of consistency and and productivity um i see james he uh from prism here and samarasat uh here so uh thanks to their contributions so yeah um good deal i love rated i cannot like this this is just one of my favorite dashboards and you said they've added an api uh well they have an api but somebody third party has added a prometheus

exporter which gives you command line access to the api so you can kind of programmatically query it much more easily i haven't had a fiddle with it but there's a a lot of rich data there which you can slice and dice in uh you know all sorts of creative ways so you know rated the front page is one thing but actually it's the underlying data set that's got all the really interesting stuff in it so uh now that it's easier to to query yeah that's great so i i have to admit uh i'm enough of a noob that i thought this was gonna be a video and i love video demos of any product like it just it takes all of the work away of like reading you know the the man page uh and just like watching

somebody else do it but uh yeah that's i love to see this example down here um i'm a big fan of rated and appreciate what they do they they were really great with client diversity by cutting this breakdown into individual clients so very cool one of my favorite people ever joe clappers have you have you ever chatted with joe i haven't no this is great though this is just like a humble genius kind of dude um but yeah joe has been working on this i want to find it joe has been there it is joe's been working on um

these low power and so for those who don't know joe is a developer for rocky pool and he's been work working on these low power staking and so this this is incredible yeah uh eight watts while sinking five watts during validation i really that's astonishing isn't it it is and he has engineered this thing so well like this is what just what joe does so i would also suggest giving this guy a follow he's he's just a gem sorry you know i'm supposed to let you

tell about the context and i just i hate it i don't know when you've got better context that uh that's good with me i didn't know that uh jcrpd was was with joe even so i learned something okay yeah um so i there's a little backstory there um i used to do what's uh what did i call it uh uh steak state of the state um i was you know it was a like a eight or ten minute thing and one of one of the state of the states was a plea for developers who could implement client diversity uh and so after one of those joe who had very little experience was like hey i'm interested in ethereum what can i do

um and so uh after some talking he got plugged in with rocket pool and he's the guy who implemented all of the four clients into rocket pool and now he eventually got hired by them uh but he just has this this deep interest in supporting diversity uh and i'm i'm a huge fan of his as well excellent hero there's another hero it's a book of heroes yeah i was reflecting earlier that um adrian is pretty much unique in this world in the sense that he i think it's correct to say he's the only current dev who's got um you know been neck deep in both

ethereum one and ethereum two development i mean he was a devon basu for a couple of years and he's been um in a tech lead for teku for uh two and a half three years now and i don't think there's anyone else involved who who can genuinely say they've been fully um dev on on both sides of the the merge as it were um meredith meredith baxter who used to be in tecu team she she had that honor but she's deserters now she's gone off to the uh cto somewhere else so i can't comfort her decision but uh sorry sorry to lose to that but so that that leaves adrian um and it means it makes him really worth

listening to because now that we've got this deep interplay between the execution and consensus layers the the insights that adrian has um are you know well informed by both sides of this and they're always consumable like i i'm always able to understand what he says even if it's complex so i do appreciate that from him yeah i don't know where we'd be uh with techy without him he is superb um i loved this tweet from pintel uh again getting back to things that fitz can understand so uh trent van epps said uh who can describe uh the kzg ceremony and what it enables

and so you know most of us morals have like zero context here and i want to read this just because it's so well worded by pentel to enable so this is what the kzg ceremony is um so to enable data availability sampling a key part of dank sharding we need to encode the data in the form of a polynomial this allows us to reconstruct all of the data if up to 50 percent of it is not available from the network perfect like i can wrap my mind around that thanks fintel yeah pinto's made a good uh i won't say career but uh has gained a reputation for um clarity of writing i think he first

came to my attention with um there was a madasha challenge to write some analyses of um how the network was going and uh um you know analyzing different aspects sort of open call from the ethereum foundation and pintel uh wrote a terrific analysis uh of that very accessible and some some really good insights so i had pleasure of meeting pintail at uh dev connect in amsterdam which which was great that sounds great yeah um an awesome opportunity to meet uh pentel so i have these people pentel adrian uh your writing joseph cook um like those people who can just convey complex concepts uh in a way that you

know the rest of us can pick up like they i feel like often purposefully avoid jargon and focus on simple wording and i that's i i really appreciate that because it's really it's really hard yeah i think we'll just get to that when we talk about the book but you know it's funny i've i have never read straight through uh but i do have like several portions of the book bookmarked now because i'm like that is my best reference for something like rewards or something like uh you know whatever it is like it i have a lot of great go-to's for that so i'm really happy with it um keep an eye

yep can't hurt as busy as ever so they're going through doing lots of interviews with um uh people from the different teams uh i saw a tweet earlier there's another one coming up um so yeah as ever just uh spend some time on the um cat herders peep meep um youtube channel uh there's just such a richness of material there from the horse's mouth i mean it's the people doing the work who are talking about their own work and uh is his quality yeah and the this is one of those under-appreciated i i think in 10 years it'll be great to go back and look at the quality of this information in it because i don't think it gets enough attention now um but i do

think it's gonna it's gonna be a great testament in time to what people were doing um yeah absolutely uh the book progresses slowly so slowly um but i uh i thought i'd just bang out a quick um section just to kind of get back into the rhythm of um writing things and so i did this fairly straightforward thing on aggregator selection i mean nothing is straightforward right but and then uh yeah just uh easing back into the routine of writing i'm i need to make a lot of progress um unless the merge is delayed for a couple of years no jokes

i i'm sure you'll do fine so ssv is hosting twitter spaces or anthony cesano is hosting yeah his second mention uh in this edition he's uh he's busy um yeah they these are nice i mean um i i haven't had the opportunity to hear tim's or stannis i uh listen to my podcast when i'm out walking i don't know how to download twitter spaces so that's i could probably work it out but if i could download it i'd listen um but uh so they uh um they did that so ssv network um they're building distributed validated technology kind of um alongside obol uh and um they're doing a series of

these i think the intention is to do one a week though i think uh steny kulatov's was was delayed by a few days uh one a week until the merge so uh depending on when merge um there should be a good good set there yeah yeah at some point they're going to say we're out of guests we need the merge which is just like it'll be a long time oh there's another guy there yeah i had the pleasure of talking to anthony on uh friday morning and um it was it was a good chat uh enjoyed that and um he led it very nicely wasn't even aware of that you see it's you snuck it in there well good i'll i will definitely check that one out today

and so capping the validator set what do you think for people who are listening i should i i one of the things i want to improve is giving context to what we're reading because a lot of people will listen to this while they're walking and not look at the screen um but yeah vitalik opened up the idea of capping the validator set and uh now that's my setup sorry yeah so this has been talked about for a while um there was a suggestion of putting in something into the altair fork that we did the alt air upgrade that we did last uh october time to limit the total validator set that's active at any

one time um and vitalik in this article sort of is he's not arguing for or against any particular position he's just sort of laying out various alternatives and their pro and con on there and there are always trade-offs right uh with with doing this but the idea is that um there's no point just allowing the valid data set to grow unboundedly at some point it it just becomes too heavy for nodes to um to compute and doesn't really add any more useful security to the network you know there comes a point when you've you've just got enough security and adding adding more just adds overhead so

um yeah there's questions about what is the right number and what are the right sort of incentive models um around you know um cap not yeah hard cap or soft cap you know do you change the incentive so people self-limit or do you um put in an in-protocol limit or you know do you um stop new validators registering or do you kick out when somebody new registers do you kick out an old one or you know so he goes across all of the uh current alternatives that are in view and um recommends one at the end but uh there's a kind of least worst option uh i i think the size of the validator set is just

well first i should confess i haven't read what vitalik wrote but i therefore i should not comment period but you know we originally had this idea that uh locking up 32 east would take it out of circulation but now we have these the proliferation of lsds and so there's there's no real consequence there's no you're not locking anything up you people still have that full liquidity um i just find that really interesting so yeah uh the dynamics will change when withdrawals are available um and he also talks in that article about potential to reduce the stake size which is kind of

counterintuitive because you think if you reduce the stake size and you you just end up with more validators and you kind of double your problem but he's got suggestions um that's kind of uh consolidate stake on validators so we don't have to have 32 weeks per validator you could have a single validator that has a hundred eth which is his weight in protocol and another validator that's got 15 heath and that's his weight in the um and and we'd have to change some of the way that the the these committees structured and so forth but it you know that that could become an interesting possibility i i want to make an embarrassing confession

so i opened this and i was like okay let's see am i going to read this and i zip through and i see an equation and i'm like nope i'm like that's not complicated this is what uh stephen hawking's publisher said to him uh way back when uh he published the uh brief history of time this his publisher said you have every equation you include is going to halve your sales so he just put he has one equation which is e equals nc squared you know the famous one that's the only one in the whole book but and i love the book i love it because yeah but uh his publisher knew what they were talking about i'm serious like i um i'm not i'm not math incompetent but i see

equations and i'm like i just you know sorry all core devs we kind of went over that i haven't seen christine's notes you're a huge fan of christine's notes i haven't i i followed them for a while but i've kind of gotten away from them i'm big fan of christine around and we had fun doing the coindesk podcast uh together and uh yeah i mean christine she was a journalist she's kind of an analyst now rather rather than journalists and she's thorough she doesn't hype things up she gets her facts straight um and uh is you know very detailed i'm i'm impressed with the level of detail she engages with in her notes they're much more detailed than mine i mean i i made

mine at the time which is uh you know as the call goes on but uh christine and tim sort of go back over the material so they can be a bit more detailed but uh you know christine really engages with with the deep detail it's um it's uh impressive stuff and she's taking pains to write a lot more about ethereum now yeah i i i really need to appreciate her more i i will be overly honest when when she was in this thing about the merge has been delayed and i'm kind of like well no the merge hasn't been delayed it just hasn't happened yet um i i sort of created a mental block and that that was a really poor choice so uh yeah i mean you know she and i don't see eye to eye on on

everything but i i think she is intellectually honest which you can't say about there's a good fraction of the media which you cannot say this about absolutely we don't have to agree on everything in order to be on the same page so yeah i i do want to get back into her stuff let's see kcg stuff merch call community five that's actually an important one merch called community five uh that's this friday um and yeah uh there's the agenda i'll try and show up if i can um and it's uh does that clash with all core devs might do that would be unusual i'm sure it might be we're moving oh it's not this week yeah no it's um

consensus implementers this week and all called those next week last one on a friday next week and then moving to thursdays after that all right uh let's see i'm curious mostly british 80s music what what are what are some i'm just give me one or two things that i would recognize here i need i want to get a i want to get a feel for what uh ben's mostly british 80s music playlists the the mostly british is because i do have some u2 on there which is irish and they'll be very upset to be called british so um perhaps you know it's kind of uh pet shop boys depeche mode all the kind of synth fault stuff alison boyer

and uh all of that uh yeah i love all that synth pop um good yeah i i was really just curious to know uh being on the wrong side of the pond if i would actually be able to connect with it so yes i can splendid i i gather uh yazoo uh was known uh over your way as yaz due to a trademark conflict but uh this is um alison moyes and vince clark's band uh which uh we knew as yazoo rather than yes very interesting yeah um and so before we kick off uh i would love to encourage everyone to go follow ben on twitter he's minion underscore xyz um always the the best stuff

oh hey there's one thing i do need to talk about out there nope we'll get it next week um yeah uh now i'm gonna leave everybody hanging until hanging a fortnight uh it has been great hanging out with you i always enjoy doing these um i am caught up on what's new in heath too yeah good stuff uh thanks for as fun as ever to uh just to review uh everything you keep me honest i that's my job all right buddy take care okay so first stop share and then end bye everyone