Sepolia merges, stETH recovering and much more - The Daily Gwei Refuel #401 - Ethereum Updates

Transcript

hello everyone and welcome to another episode of the delegate refuel worry cap the latest news and ethereum ecosystem i'm your host anthony designer and today 7th of july 2022 all right everyone let's get into it so sopolia merged successfully ah it's so exciting right like so very happy that it all went through well i'm gonna spend probably you know half of today's episode talking about the sapolia merge and just merge related things in general um but basically uh thanks to everyone for watching the stream if you kind of like watched it live or kind of re-watched it it was a really really fun stream a lot uh more i guess like uh organized than the robston stream because of the fact that we didn't have any um kind of like issues with hitting ttd uh because of the fact that the hash power was there it was right on time

basically of when one of when the accordives wanted to happen and it happened and also sopolio is a permissioned test net which means that uh not anyone can just uh sorry the beacon chinese is a permissions test net so not anyone can kind of like jump on it's kind of like whitelisted and things like that so it was a bit smoother of a kind of like transition there as well but yeah i mean the train finalized like there was initially post merge a few kind of like issues around about 25 to 30 of the network not proposing blocks and that turned out due to to be due to a misconfiguration error as usual with these things what's funny that for the first few test nets it's really just been configuration errors it hasn't been anything kind of like merge related and they were fixed quite

quickly and then we're back up to well we're basically back at 95 participation now 94 now the reason why it's not close to 100 i mean one it's a test net so you know if you're kind of like on a test net even in a permission test net you have less of uh incentive much less of an incentive to keep your nodes online because there's no money on the line right it's just test net eth whereas on mainnet you know the validator participation is at 99.7 percent so there's only 0.3 of validators that are offline so obviously a very large kind of like chunk chunk there uh but uh but essentially if you kind of look at look at this now 94.5 and the missing validators as well rather than kind of like just the ones that uh that are offline due to this

just being a test net there's also a bunch of keys to validators that were lost i think um when this was spun up which means that those validators are basically dead validators at this point but they'll get kicked out eventually of kind of like the validator set there and then maybe we'll see this creep up to something like 90 uh 99 as well so i do want to kind of like break down a little bit kind of uh what happened there are a couple of threads here one from terrence uh who who's um one of the founders of pry labs here so his summary was basically merged it was a success 25 35 of our validators went offline uh because of this misconfiguration uh uh issue there was no client bug these hiccups will obviously not delay the merge and this is what testnets are for to basically

show these kind of like edge cases that i guess like people don't kind of like account for now the the kind of like issue basically was with the configuration is that i think there was like a ttd override um and kind of like some uh some kind of like other things got to do with that uh i think no no i think it's only the ttd override for this particular test that there have been other configuration errors but basically the ttz override kind of like happened because this is a test net this is probably not something that will be done on mainnet uh when you kind of like look at that sort of stuff look at it from that perspective this is really just a test net thing so it's kind of funny how that works where we're finding these things on test net that aren't really going to

be an issue on main net like we want to find things on test net that would be an issue on may on on main net but we're not finding any of that because it seems at this point like we're literally just going through the motions like the merge transition uh you know i think the core devs and researchers pretty much are happy with where it is today they know what they're doing you know we kind of like have it have it down to like a a nice exact science at this point and there's only one more public test net left right there's girl girly left which i think should be scheduled on the not the core dev call this friday but the core dev call after that so the the one on the 21st of july that's when they'll schedule the the girly uh test net merge

and then it's main net and then we're done guys like that's how close we are right now we have one more public test net left then we have main net now shadow forks will keep happening every week uh parry confirmed that last night on the stream so that's all going going well as as well there doesn't seem to be any issues with those things besides the configuration issues that always seem to kind of like pop up here but as i said they're not merge related and what i mean by merge related is that we're looking for kind of like something that goes wrong during the merge transition like a consensus breaking bug or something with the clients breaking because they just weren't talking to each other due to some other kind of like merge related thing uh but it's just been these configuration things which is obviously hugely positive now

this brings me to one other thing that i want to talk about which is timelines i did mention in the daily discord yesterday that i've revised my timeline a bit here uh due to kind of like this configuration issue now as i said it's not merge related it's not something that is going that kind of like is related to the merge transition itself it's not something that would break the merge but i think that my read at least is is that core developers now want to give people even more time to make sure that their configurations are correct and that everything's up to date and everything's kind of like smoothly going to smoothly transition over because on mainnet it's very different to the testnets and there are things you can't test for on mainnet ethereum uh that sorry on test nets that exist on mainland ethereum one of those things is

all the infrastructure that exists for mainnet ethereum all of the exchanges the service services all the middleware everything that sits around ethereum mainnet that just simply doesn't really exist for test nets and obviously the user side as well and users actually kind of like interacting with the chain and using the chain and putting transactions on the chain so there are just some things you can't test for even with the shadow forks right like that it's just there the shadow forks don't have that infrastructure that the um that mainnet has and the infrastructure needs to be updated like if the infrastructure providers are running their own nodes and they don't upgrade those nodes or they don't have the right configuration set then they'll get kicked off the network and they'll they'll need to kind of rectify that so you know i expect this to kind of like happen anyway i

don't expect look look i i think even if this kind of like happened on main net where 25 to 30 of the validators uh went offline or were kind of like offline and then came back online a couple of hours later the biggest impact that would be felt is because of the fact that there's missed blocks the user experience would degrade for maybe two to three hours where there would be long block times essentially because there would be like say two three missed blocks in a row which is 36 seconds and then you'll get confirmed on the fourth block but then the transactions would build up and everyone will be trying to outbid each other and it could get like i mean depending on how many blocks are missed it could get like a bit nasty with mav and things like that but generally people would be much quicker

to get their validators back online as well because of the fact that they're losing eighth right there it's kind of like a monetary incentive there i even know that it wouldn't get slashed for something like this they definitely would have like that inactivity leak and if there's a they're a big staking operation that can add up very quickly obviously so uh because of that i think that they're going to want to give people more time for girly to to prepare for girly which i expect to at this point in time happen about early mid august i i i think mid august might be more likely because if it gets announced on the core dev call on the 21st of july they may want to give people at least two weeks and probably up to play it on the safe side three weeks so that would

bring us to around mid-august now in terms of mainnet i would i would imagine that they would love to give people at least a month right at least a month uh if not six weeks so that would bring us to late september instead of mid-september it could still happen in mid-september just depends how fast girly gets done but i think that they just want to give people enough time maybe an extra couple of weeks then that otherwise would have given people to make sure that everyone updates because it's better to just kind of run through a transition as smoothly as possible rather than having to fix things after the fact but again it's not a huge deal like i mean the beacon chain has been built to be very very resilient against these sorts of things the chain was still finalizing and the reason why

it was still finalizing is because more than 66 of the network was still online more than two-thirds of the network was still online and proposing and as i mentioned the biggest issue was that kind of like degrade degrade degrading of um of of uh block proposals which is degrading the user experience so that would obviously not be fun it would be a few hours of pain there'd be a lot of fun as well because for people who don't understand how the peaking chain works they'll be like oh my god like 33 of the network went offline or 30 percent of the network went offline beacon chinese doomed all these sorts of stuff when in reality that's not how it works and i know that you um people that listen to the and watch the refuel no that's not how it works um but yeah on the timeline

front of things yes there's only girly left until mainnet but i just feel like they want to give people more time and there's also the execution layer and consensus layer forks balance tricks for consensus layer and paris execution layer which will get those two layers ready for the merge uh basically they're gonna happen to very little fanfare it's basically the same thing as a difficulty bomb delay like it's really nothing to to write home about but once those are in then we can set like obviously the ttd and then the merge main that merge transition will happen at that ttd like it has happened for the test nets now leading up to the merge we are expecting a bit of kind of like miners to fall off i mean a fair bit of miners have already fallen off the network due to the fact that the eighth price went down so

you've probably noticed a bit of hash going offline but i feel like come august there's gonna the smarterminders the ones who actually think the mergers happening are gonna probably uh begin kind of like uh i don't know either offloading their gpus or setting up future sales of their gpus maybe they wanna mine you know right up until uh right up until kind of like the merge happens but definitely would expect to see a hashrate drop there and this is what ttd was created for by the way like for these kind of uh these these things that are gonna happen with miners it was created to kind of prevent any kind of like minor shenanigans like attacking the network or anything like that but that's basically in my mind very very very low chance of happening i don't think it's really going to happen at all

but the i guess like the the other thing is the fact that miners start going offline before the actual transition happens and we want to make sure that the network can run through the transition and i don't think that's going to be an issue either i think there are definitely going to be people mining right up until the merge and because of ttd it kind of like lessens the blow there as well uh even if kind of like miners go offline and i think as well miners some gpu miners are going to get kicked off the network as well because of the fact that the dag which is the um the the basically mechanism within eth hash is what uh was what kind of like um ethereum uses for mining i think it's going to kick out four or six gigabyte cards uh from the network so

anyone who has a gpu with that much uh vram will be kicked off and they'll have to mine with uh kind of like eight gig 10 gig 12 gigabyte cards from there so there are there is going to be a drop in hash rate but don't be alarmed about it that was expected that's the reason ttd exists um but yeah the main net merge transition i'm expecting probably late september at this stage and look depending on like how much time the core devs feel like they want to give people and give kind of like all this infrastructure and this ecosystem uh kind of like people to update their nodes it could even push into october but i would say that's a low chance like i think people just want they want to get it done but it's more about kind of like getting it done

as as safely and as cautiously as possible and just being kind of like a being on the on the safe side of things right like making sure people have kind of like upgraded in time making sure they're not left behind but at the end of the day the test net recovered very very quickly here like within a two to three hours you could imagine main nets uh recovering even quicker than that because of the fact that there's real money on the line and i would assume that most people are going to make sure that their um their nodes are up to that and that everything's up to date but i would expect the corder is probably wanting to give at least a month's notice from i guess like uh when they announce the date till the date actually happens or when they announce ttd to when it

actually happens uh if not six weeks which if girly happens early to mid august and then we get a main net uh kind of uh merge uh date shortly after that i would say it would push into to kind of like late september there but we're pretty much there like as i've said before we can see the light at the end of the tunnel we're just running through the motions one more public test net left uh got the paris and balatrix forks to happen which again are nothing special they'll happen with with little fanfare and then the main net merge uh happens and like i just i was kind of looking at this tweet actually this is this is something that i want to talk about this tweet from um trent van gaapsia and he he basically kind of like showed all the testing all kind of like the major testing

efforts in order for the merge so we started with rayanism and now i spoke about this a long time ago i'd actually forgotten about rainism this was born out of a hackathon i'm i think proto lambda and a few other people kind of created this and this was a way to initially test the merge transition then we had amphora we had kansugi kiln we've had the girly shadow forks one to four main net shadow forks one to eight merged devnets one to six robs then sapolya and now i've only got two things left girly and maynet plus of course the shadow forks that are gonna be happening every week uh but that's kind of like the regular cadence right up until the merge transition happens there but like i find this just really really amazing because we've tested the merge so much at this point

guys like look how much testing has been done and not just the same testing different types of testing and testing on new test nets testing on existing test nets doing shadow forks which are basically a new thing for the merge transition we didn't do shadow forks before the merge before we started testing the merge which is actually quite wild because they're really really cool things um to do and they gave us a lot of insight uh but like there's two items left on this list now girly and main net i know i'm i'm kind of like you know repeating myself here but like for me for someone who's like spends his whole life in ethereum and following this so closely for so long now like when i got into a theorem in early 2017 you know proof of stake was one of the main reasons i got

into it and was so excited about it because i was like oh my god like if this works this would be really awesome for this this and this reason um and then i watched it evolve every time i watched all the trials and tribulations in real time i've been a part of every single major event that's happened uh with with regards to proof of stake and we are literally just a few months away from the main net merge happening from ethereum being a full proof of stake network and we can uh kind of like move past all of the shitty narratives that surround proof of work and i've mentioned before how i don't have a strong opinion on proof of work uh if it's good or bad i think that if people want to fight the proof-of-work fight in in kind of like the mainstream they

they're more than welcome to i encourage them to do that but i'm just very glad that ethereum won't have to deal with that because ethereum has enough things to fight like i mean come on guys like we don't need that one just hanging over us we i'm just glad we're moving to proof of stake for that reason but also for every other reason that that you guys know by now uh but yeah so close i'll keep you guys updated on the timelines here but as i said i think girly will be early to mid august at this point it's definitely not going to happen in july like i just don't see them it's already july 7th as i said they're not actually talking about it until july 21st because the tour dev called tomorrow they'll be talking about the sapolia merge and lessons learned from that and all kind

of like related stuff to that i think dang sharding's actually getting spoken about as well so i expect girlie to get a date on the 21st of july which will put it early to mid august in my in my eyes probably closer to mid august and then may net probably late september or something like that which again is still very close to july july august september it's like three months like three and a half months or something uh it's gonna go quicker than we than we kind of expect i'm looking forward to it obviously and i can't wait to to kind of like see it there but i think that is it for most of the stuff that i want to talk about with regards to sapolya there is the uh song that jonathan mann did uh the song of day man on um on the sepolium livestream that we did

he basically uploaded it to twitter here it's about eight minutes and 30 seconds all the lyrics for the song came from the daily gray discord channel which was really really cool and he added some music to this as well so if you want to go listen to this you can there's also an nft you can bid on i'll link this in the youtube description below for you to do so all right so speaking of proof of stake we now have 13 million each staked in the proof of stake deposit contract of course this is mainnet this is not the test nets and this is still worth almost 15 billion dollars like i know the east price has come down a lot it was worth a lot more once upon a time but the thing is is that you know we all expect the eighth price to go back up so as the you

know more and more eighth gets staked it can't actually be removed for at least six to twelve months after the merge because withdrawals aren't enabled so this number could get quite quite wild we could actually get to a point where we have a hundred billion dollars worth of each securing the beacon chain before withdrawals are enabled if the eighth price uh you know goes crazy again i mean you can imagine us getting to 20 million each staked and then the eighth price is like 5k i mean the price was almost 5k uh as his last all-time high um and then we would have a hundred billion dollars worth of eighth state obviously like i got no idea what the price is going to do um but you know you could imagine a world where that kind of happens which would be quite quite cool to see but just awesome to see that

there's still people kind of like staking still people going into staking i mean if i go to the kind of like beacon chain website here we can actually see uh so i've got the sepala beacon chain side up there and i've got the main net up here you can see you know we're at 400 and almost 106 000 validators uh there's zero pending validators right now uh but you know there's um 13 million each staked there and as i said like 99 you know look 99.6 99.5 percent participation here which is which is obviously what we want to see and that will kind of like consist after the merge transition as well you know even if there's configuration issues they'll get fixed and then we'll just go straight back to this year so that's that's basically what maine that's gonna be anyway uh and speaking of kind of

like uh i guess like staked ethio there is a um a great a great kind of like tweet here from uh from michael i believe his name is where he shows the st price has been recovering since mid-june and the discount to eat is now less than three percent so you can see it on the chart here what the discount got to i should have bought some like i saw it get to like not 0.93 ish and i'm like ah like i should buy some here because there's so many distressed sellers right now i don't think it's going to go much lower than here but i didn't i could have got like you know a free bit of eighth there but i didn't but that's okay i still bought eighth you know i feel like i'm out of the game um but yeah and then

michael continues by saying large selling stopped entirely after celsius and the st liquidations interestingly no large buyers either just smaller ones bringing up the price increment incrementally so it's pretty cool to see that the smaller buyers that might mean that there's just a lot of enthusiasts a lot of retail investors who saw this trade and they're like well i'm gonna take this like yes okay it's got some risk attached to it but it's really relatively low risk and you know that you can redeem st for one eighth uh once the kind of um the withdrawals are enabled on the beacon chain which should happen yeah as i said six to 12 months after the merge and if you're happy waiting if you're already like a kind of um uh someone who's long eight someone who's bullish on eight for the long term then this was

a great deal if you if you're kind of like happy taking on that that little bit extra risk and you're happy kind of like waiting it out until it matures at beacon chain withdrawal there so cool to see that this is recovering of course obviously some users got hurt from this especially if they panic sold it thinking that it was going to death spiral like lunar and ust which is something that a lot of people were saying they thought like because we just had the terror collapse everyone was speculating that everything that was quote unquote pegged to something was going to death spiral we had people fighting tether with people fighting usdc where people going after all the other algo stables which to be fair you know most of them do go to zero well pretty much all of them go to zero actually i shouldn't say most pretty

much all of them have gone to zero um but as i've explained before st8 is an r8 and other kind of like liquid staking tokens are completely different they're collateralized by the eth in the beacon chain like just because you can't withdraw it now doesn't mean you can't in the future it's not under collateralized or uncollateralized like some of these or all of the algorithmic stable coins out there so just wanted to kind of i guess put that out there and yeah cool to see this recovering hopefully the recovery continues all right so i mentioned this ama that the ethereum foundation research team is hosting it's happening i mean it's probably already live by the time you're watching this uh so definitely go check it out if you haven't yet i'm sure there'll be answers up there already i'm

looking forward to this as i said yesterday these are always great there's always very very informative and pretty much i think everyone from the ethereum foundation research team is going to be there all the people that you hear me talk about a lot on the refuel such as justin drake and dan cradfeist here um of course vitalik's going to be in there i mean oh wow dan great actually has a tweet here listing them all danny ryan um xiaoweiwang uh who's this asn who i actually am not following that must be a newer research a new ethereum foundation researcher um and and ansgar here as well who i believe is a new researcher too but yeah the who's who of ethereum kind of foundation researchers it's definitely going to be a good ama you should uh definitely go check it out i'll link it in the new subscription

again here all right so uh this account dino eggs on twitter has put together a last little uh image uh by and he captioned it with my take on polygon security spectrum and where each solution lands from standalone chains to fully secured roll-ups now the reason why i think this is a really good graphic is because i think people still get confused about polygons suite of applications or suite of products i should say and where they all fit in so you can see on the left-hand side here and my camera's covering it but on the far left it says not secured on the horizontal access yard and it's got po uh uh not secured by ethereum as you can see on the bottom secured by ethereum they've got polygon edge uh with like right at the kind of like not secured

side of things because polygon edge doesn't talk to ethereum um at all besides kind of like i guess the bridging and people can post data there if they're in l2 but it is not secured by ethereum's validators or miners then you've got the polygon pos chain which is a bit further on the spectrum that the polygon post chain has its own validator set but it also check points to ethereum so you know if you had to you could reconstruct it from from the checkpoint so there it is it is secured by ethereum at least a little bit there then you can move uh further towards the right of the secured by ethereum uh horizontal axis here and you can see valydiums and and falling to that is volitions as well and polygonal veil uh sits there which is the data availability layer oh so i got confused between edge and veil

edges the sdk uh i believe uh polygonavale is a data availability layer um that's where data gets thought on a veil and the proof gets stored on ethereum so it's like a half secured thing i mean don't quote me on saying half secured because it it if you're gonna go into like really deep semantics into this it gets it gets funny and researchers are going to yell at me for saying that so don't quote me on that but the i think i still think this does this paints a good picture and illustrates uh kind of like how it works so probably gonna veil obviously data availability on polygonal veil proof for um kind of validiums and volitions on ethereum uh and then the volitions uh that kind of like fall into that can also choose to store their data and their proofs on ethereum which is a

fully secured chain fully secured l2 chain as we know and they're called roll-ups and this is where four of um polygons products fit into polygon hermes polygon midan polygon nightfall and polygon zero so that's the spectrum for you guys to kind of like look at and i think the reason why this is so great is because you can quickly use this to to show people hey you know polygon is a polygon does have layer two products but it's not technically the pos chain it's actually these things on the on the right hand side and then you can educate them about uh kind of um data availability and data and proofs and things like that but it's really cool to see that actually most of polygons products fall on the fully secured by ethereum side now obviously not none of them are alive in any real

capacity yet i think hermes is live on um kind of like mainnet same with nightfall but they're not obviously you no no they're not used nearly as much as the polygon pos chinese but the plan is obviously in the future for them to be to be used as much as the pos chain if not more especially as as we add more data availability to ethereum layer one so that they can have cheaper transactions to be secured by ethereum there so yeah just wanted to kind of highlight this because i thought it was a really really cool graphic so speaking of polygons uh kind of stark solutions here i'll layer two solutions here uh brendan who is working on polygon zero put together a great tweet thread where he starts off by saying the recursive starks pioneered by polygon let us paralyze proof generation and

reduce latency but what's really cool is that they allow us to exploit a space-time trade-off this lets us do crazy things like proving 100 hashes per second on a laptop now this is getting a little bit into the weeds well a lot into the weeds of the cryptography side of uh kind of roll-ups here and of things like starks it's a short thread if you want to kind of read it but i think from my read what he is saying here is that uh paralyzing kind of like that proof generation reduce reducing latency allows us to basically do this proof generation on a laptop on a lower powered device which is what we want to work towards because as i mentioned yesterday i think the day before there's going to be asics for that prove these sorts of things so

sorry that prove starks and snarks and all these zk related um cryptography here and if we can get like really really efficient assets we can be proving a ton of these things and and kind of like i get zk rollups to be extremely cheap on the proving side of things and essentially get them to the point where we can scale like up to i don't know hundreds of thousands millions of transactions per second that's where the craziness kicks in and i've talked about this before how i think that we're gonna get there it's gonna take a little while uh probably a longer while than we would like just because this stuff is bleeding edge it's using our new cryptography obviously whenever you do that you want to be very careful with that but we're there the path is there we know what we where we

want to get to and there's so many teams working on this obviously not just polygon there's so many teams working on uh stark snarks zika roll ups all these sorts of stuff uh obviously starkware it's in their name right uh and then you have like uh scroll when you have uh zk sync and you have a bunch of others out there i'm sure arbitrary and optimism are also looking into this sorts of stuff too so yeah just really really cool to see that uh you know we're already getting to a point where this this kind of stuff goes beyond theory and starts going kind of like going into practice which is definitely what we want to see because that's what users care about at the end of the day right all right so just wanted to quickly shout out again the daily gray tv account here so this twitter account is

maintained by zk ben who a lot of you may have seen if you're in the discord channel he's quite active in there but for those of you who don't know the daily gray underscore tv account on twitter basically shares short clips from the refuel about 30 second clips and also extra information uh from kind of like the refuel as well so kind of like links and things like that so definitely go follow this twitter account maybe you are kind of like missed the refuel maybe you want to share one of the kind of like things that i'm talking about on the refill to your friends in in a short bite-sized clips well this is what you can do on uh with these kind of like daily grade tv clips here this has been going for i think since june now so a couple uh i

guess like six weeks almost now or five or six weeks uh nzk ben has been doing an amazing job so obviously very big shout out to him there uh but yeah this was just put together by him i didn't even ask him to do this or anything i didn't ask anyone to do this he just went hey i want to do this he asked me and i'm like yeah sure man like go ahead this would be awesome so if you're not following that account it's linked in the youtube description for you to go follow it please do if you even if you just even if you're not even interested in the clips you just want to support the the account obviously if you watch the refill every day you probably don't need to watch the clips but if you just want to support the account you can

follow it get that follow account up so more people find it more people find the delight ecosystem and we can grow as a community as well there but i think on that note that's going to be it for today so thank you everyone for listening and watching be sure to subscribe to the channel if you haven't yet give the video a thumbs up subscribe to the newsletter join the discord channel and i'll catch you all tomorrow thanks everyone