What's new in Eth2 Number 91 - 8 April 2022 with Ben Edgington and superphiz

Transcript

hey everyone i'm superfiz from the east staker community and today is it's not april 8th it's april 11th 2022. um i'm here with my good friend ben edgington how are you ben hey fears good to see you again it's been a little while uh yeah we're away on vacation and we skipped the last one of these but i saw you at the um gitpo app event on saturday so that was nice yeah git pop is it's really awesome i'm really uh stoked for colfax and for for the poep community that it launched so uh i i hated to have missed how do you say two weeks ago in in in britain

so we i kind of like we bantered about catching up but then it just didn't work out so i guess we're gonna pick up where we left off um and did you have any highlights from the previous um uh issue that you want to bring up now um the the main thing was uh writing about i'm rather embarrassed to say the uh teku incident on on mainnet so um we we didn't exactly break beacon james maynet but we um take you had a little wobble and managed to uh bring down about ten percent of the network for about an hour um and it's all fixed and it's all good and you know we we

found out what the root cause was and and fixed everything up but i felt i owed everybody a good uh detailed explanation of what went on and we published an instant report so uh i i dove into that fairly fairly deeply that was the main thing i i read that on the beach actually it was worth the time a main net incident yeah so um well so you know it's it's good to remember that uh client implementations will occasionally struggle it's why we have this client diversity um and as long as we are well distributed between the clients any kind of wobble like that is easily recoverable regardless of of one tiny slip up

congratulations for all of the great success you've had with deku so let's focus on the big picture right indeed yeah and it it is i mean coming to the client diversity it you know there are there are two ways looking at it i mean most chains say having more than one client makes you more vulnerable because you're more likely to get a bug you've got you know x times the number of code bases therefore you're more likely to get a bug and then you know more likely something goes uh goes wrong um but the way we tend to look at it in ethereum is about resilience right if you get a bug in one then the others carry the chain and that's exactly what we saw uh happen uh in in that instance that's exactly how it should be techie's less than 30 percent of the network so

you know everything just uh um breezed on you know techy users were inconvenienced for a short while but otherwise the chain was perfectly fine so you know this is this is exactly as you say this is why diversity matters this is the resiliency it brings and that's why we're so big on it in in ethereum yeah and i look forward to talking more about that uh as we get into client diversity in this issue so we're gonna go through your uh what's new in eth2 for april eighth um i did skim it but i didn't i'm not really familiar with everything and i'm still catching up from uh from time off so um

all chord devs update that i i the the naming or the numbering system for that isn't uh i mean it looks binary doesn't it um yeah i can hardly believe today is that call number two [Laughter] these are great this uh so i was writing what's new um like four weeks ago and uh or whenever it was no just over two weeks ago the previous edition and um it gets to sort of quite late here because i procrastinate all day and then i kind of write it frantically in the evening and then i lose track of twitter and then um i'd as soon as i'd kind of press publish uh i noticed i caught up on twitter and uh and there was tim's right up and it

just missed getting in so i felt bad about that because it's a really really good um summary of where everything is on uh not only merge things but also the other topics the other candidates for inclusion in a post-merge fork so um yeah so it was my duty to get it into the next edition little late but uh uh it's all still completely relevant completely valid yeah it's great actually i this is my first glance at it um and now i'm absolutely committed to reading it there's i can already see a bunch of um really valuable stuff in there so um thanks tim yeah tim's doing a super job of communicating uh at the ef i mean he's uh i

i always um when he left consensus because we're working fairly closely together at consensus when he left i i just thought what's the sort of tribute how do i feel about tim well you know when i when i think to myself when i reflect on my own performance how could i be better at what i do better at my job better it's you know life in general my my general reflection is yeah i need to be more like tim he's just doing a super job sorry to lose him to the ef but but actually you know he's he's doing such a great uh role there that that i can't be too uh too unhappy you could uh you could knit yourself a bracelet that says wwtd uh what would tim do and anytime you know

your situation you can just say what would tim do right now um yeah he's he's a great guy i don't know him that well um i know him a little bit but he's he he is just killing it in uh in the ecosystem so very happy about yeah yeah totally yeah another tim product yeah so this came out of a thread on twitter which was um somebody very very frankly and openly saying i've got no idea what's going on which is great i love it when when people you know most of us pretend we've got some clues as to what's happening but it is confusing and it changes and and information comes in drips and drabs from all over the place it's it's hard to make sense of so when somebody

actually you know um points out the emperor has no clothes no it's not quite the not quite the analogy but when somebody speaks up says look this is too confusing i i love it and so tim came in and uh answered had a thread answering the questions and um and then this uh faq came out of it this is an extremely concise um kind of status report yeah so that's fantastic um you know just even forget people who are starting from zero and catching up people who just get a few weeks behind like it's easy to not have any idea where we are um so yeah i was off twitter for two days last week and you know coming back i felt like i

had no idea what was going on yeah okay beacon chain it's um prism super majority uh is now less than 67 uh my favorite current source for that is uh actually sigma prime's block print as presented by clientdiversity.org i feel i owe you an apology about this i was a little negative you were you know this is the most important news of the of the year um and i was a bit bit sort of down on that rained on the parade a little um in in my comment there and uh but actually i wasn't up to date with the latest stats and so what you just showed me from sigma prime

uh is surprised me um that you know we've got uh a much more even distribution prism at just over 50 percent is astonishing from where we were just a few weeks ago and you know teku lighthouse are sort of approaching the kind of optimal 25 ish sort of uh area which is which is superb i mean teku is uh almost doubled in the last few months and yeah i'd be great see numbers getting more attraction it's uh you know got a got a bit uh going there but it is um is a great little client so uh um yeah i i'm more positive than i came across when i wrote that and that's okay like i um for me i you know i thought of it

like a log jam and for so long we were absolutely stuck 67 nothing changing for months and months um and so this this uh you know the progress probably took place over several weeks but our our viewing is delayed because it's based on block proposals um but knowing that knowing that coinbase and kraken have reduced actually there's a better uh reduced their dependence on on prism and also knowing you know we're lydo is onboarding a ton of

validators right now and and they are doing that with a focus on client diversity and so while i i have concerns about the overall governance growth of lido i'm really delighted that um that they're acting as good stewards right now with client diversity so some of those things are great to see yeah yeah and i think uh your advocacy has been relentless on this and i've uh often been uh impressed just you just keep coming back and not letting this this issue go and i i genuinely think it's made a difference i mean it has made the difference uh having you know just making noise about this i i think really does help thank you and i i'm proud of myself because i'm i'm a

quitter by nature and the fact that i've been able to bulldog this until it's it's i've seen it through as my therapist would say i've leaned in uh so i i'm really proud of that so it worked out well yeah yeah and you know we're still not quite there i mean still over 50 for one client is is not completely comfortable but uh i'm encouraged that we're moving in the right direction absolutely yeah you know it depends where you're coming from uh coming from that super majority this looks great uh if we were coming from the other direction it would be an absolute terror panic but knowing that this is improving over time

and knowing that the message is out i actually feel really good about this and i look forward to seeing a bunch of greens or greens and yellows here in the next few weeks so yeah but we'll move on thank you merge testing there was a big tweet about that this morning that i know you weren't able to cover here um yeah yeah so this is the um mainnet shadow fork the first one happened today so we should probably talk about what this shadow fork thing as far as i know it was more or less invented for for the merge so marius from the um ethereum foundation he's one of the guest developers was credited with um inventing the

shadow fork so basically um you're running ethereum main net proof of work um and you sync up a node that supports the merge so an ether one node a geth node or a or a base or whatever um but you have the extra merge code in in in that node um so you sync up the main net so you've got the full mainnet state and you're processing all the mainnet transactions and then you simulate the merge on the nodes that you control so you you sync up a dozen of them or whatever and they are your your um merge test net so at some point you're um you've got the shadow of the main net which is receiving all the mainnet

transactions and then it starts producing blocks by proof of stake and it sort of diverges from the main net but it has all the state and it processes the mainnet transactions uh and it's a brilliant test it is is effectively doing the merge um it is exactly what will happen when we do it for real with the full state and and the full cut over um at a particular difficulty height so yeah yeah it's a great test um this is the very first mainnet we've done a few uh girly shadow forks we've done i think uh three four something like that um when i say we i i wasn't involved at all um the inclusive us yeah basically barry and all of us barriers from the ethereum

foundation uh did this and um so it is a bunch on the girly network that's good but not quite such a test doing it on mainnet is genuinely a a full test of the the merge um and uh there was a tweet about five ten minutes ago i think from maris saying that it had gone uh superbly that one may not shadow fork was a huge success uh a couple of little gremlins never mind and basically stopped um fixes being deployed so you know so the caveat you know the the other side of it is um we focus on what's good because you know we're optimists and we love that but actually

we do need to focus on what's bad it does need to go perfectly we need we'll do this a few more times we will not do the actual merge until we've done this flawlessly for a few times so um but it's yeah it's good we have that whole you know it's great to find bugs let's do this and find bugs uh while we're all secretly hoping that everything goes perfect every time and then we find a bug and we say this is great we're so happy we found this bug and everyone's like oh crap oh yeah so april 29th is is inflating as a big day on awkward devs yeah and i'm i'm hope i'm right about this uh it's i it

was mentioned on the last awkward devs call as the the date so two after that one one month after that that call it was i said we will make a decision so there are various things that have to be decided so uh forking test nets are we ready to fork the actual you know like uh ring b and and all of those sapolia test nets um are the clients ready for that so that that's effectively the go no-go decision that would be made at that meeting but it it it is um implicitly a go-no-go on you know are we going to try to do the merge in time or are we going to delay it because if they delay at that point if we delay forking

the test nets then we're not going to be able to get it done before the difficulty bomb and we have to delay that so so it's a big it's a big call it's a big decision um at that point i think it's it's important for people to uh i think the term is to still themselves i don't know if that's steel or s-t-i-l-l themselves we as as end users as as eager as as we are for this to happen we have to realize that um core developers are not going to proceed until things are perfect uh and so the actual likelihood of um of this going through is is probably uh a little lower than than you know

the optimists feel excited about yeah and there are people on both sides and and the other thing is we're not doing this in isolation right when we put up the beacon chain that was pretty much in isolation we could do that as a bunch of dev teams without you know touching other parts of the ecosystem at all whereas for the merge to go well it's not only us that have to be ready is everybody has to be ready anyone who's running a node is going to have an impact from this you know whether their exchange or their infuria or um other um dap um operator or you know somebody a a home running uh running a node or a staker you know every single one of the these

entities uh uh or block explorer you has to upgrade their infrastructure and it's not a simple upgrade it's not upgrading the software on one client it's putting down a whole new piece of infrastructure a whole new client uh doing all the the networking and all of that that stuff and these people are not just doing ethereum you know for the most part the average exchange is probably sporting 20 30 chains you know this is just one among many for them they need to schedule the engineering resources and blah blah blah blah blah so it's not like we say is devs okay let's do tomorrow um there has to be a whole we have to allow um a good lead time with the appropriate support

and the environment they can test in and stage it and and and do everything so that everybody in the entire ecosystem is ready uh to do it because otherwise it's going to be a disaster as an eth maxi i'm like why are exchanges even supporting other coins at this point like why bother like yeah i i wanted to say though um the the kiln launch actually to me seemed like a really great wake-up call because it seems like testing has kind of quadrupled since kiln um and it seems like there's a real passion to kind of get caught up and realign do you perceive

yeah i think so i i mean effectively it's the first serious test net that's you know um people who who are you know using the the the chain professionally should rely on and uh you know invest in running nodes on kiln and getting their infrastructure ready it is um the closest to a production test note that that we're going to have um before the merge so that's the one if you are delaying getting going i would encourage you delay no further this this is this is a chance yeah it's coming uh and you know i i i see this optimism for july and i i'm so hopeful that uh it does pan out um

and and not in a rushed way but in a you know we've done six tests in a row everything has gone flawlessly uh let's take it to test nets i i hope that that yeah it's just three months away three months can go in a blink of an eye um and there's uncertainty about the difficulty bomb as well um mining hash rate is uh high at the moment um all-time high i believe which brings it forward right the more um the more difficult he goes up the closer the bomb comes so uh i think i think that's right so you know if the miners keep crunching away then then it will will become a problem sooner uh and we we won't have any option but to delay it so

we'll see slide over here and hit ctrl c on my miner real fast um i'm not mining except sapolia alex stokes did peep and eat yeah i don't think that's up yeah it may have gone up since i i wrote but i'm sure that'll be great um so yeah they record uh then puja puts up the um recordings a few days later um so uh worth checking the eth cat herders youtube channel there's such a wealth of great stuff there i eat cat herders peep and eep series is one of the most underrated things in our in our ecosystem and their videos

uh you know get woefully few watches and there are so they've done like a hundred um talks about different eips uh in ethereum and presentations and some are short and some are long and some are deep and some are you know not so deep um but there's immense wealth of stuff there presented by the people who are doing the work or writing the eips and so on uh it's brilliant it's it's funny because the the thing that we've kind of learned from eatstaker is you know we're working to target the audience that is ready to listen um and sometimes like if you aim too high if you're too scholarly um then people tune out and and we really figure that out by doing this um

uh how to get started staking with a nook uh with our intern mr mojo and he was as goofy as can be and it was just so wildly popular and so we're like okay we we have to we have to talk where people are and not you know um but yeah and i don't want to take anything away the eve cat herders are fantastic i love their content it is so high quality um i i worry that you know the general public may have difficulty following it uh but it's it's extremely valuable yeah it's fair point it it probably assumes a certain level of knowledge and it's from given from a devs point of view so uh yes it's it's not entry level but uh

but there are some there are some good presenters around it's worth keeping an eye on oh yeah absolutely you know um it's a little bit it's just like twitter followers like would you rather have 10 twitter followers who each you know are know everything about the subject area or would you rather have a hundred thousand who have no idea what's going on and it turns out like having it you know targeting that correct audience is really valuable so i i think eth cat herders really hits that mark in their way uh so that's fantastic ah credential migration um yeah yeah yeah this is interesting one so

um there's a long story behind this and you know talking of coming from the devs point of view um you know there was a plan to migrate everybody's ethereum account to a new kind of account where you would use your bls key to um uh to operate your account instead of your you know current key and so forth that that's long dead we're using normal ethereum accounts you know future of ethereum carry on using metamask and your ledger and all the usual stuff and signature and all of that but there was this sort of overlap period where withdrawal credentials were bls keys so there were this new signing form that we have and you made a commitment to this key so if you staked

early in the life of the beacon chain you will have a 0x00 type key so um the first version which uh you you had to register a commitment to this key um if you state later on you're probably using a normal ethereum address as your withdrawal credentials and you'll have specified a 0x01 key and it will just be a an ethereum account so because the first type is the zero type no longer points to anything that can resembles an account it's just a assigning key we we need you know some stage if people want to do withdrawal we've got to get into the system the account the

actual eth1 account they want to withdraw into so there needs to be a migration from the zero key to a zero one key uh a process of that and especially if you want to take advantage of this automatic sweeping of your rewards into your account every six weeks or so then that will only apply to the the new style the xerox zero one keys does that make any sense at all yeah it's well so i have to acknowledge that is completely new to me um so i'm aware of 0x01 and i'm aware of 0x00 but i didn't know that the deposit system had changed so that the ox01 key is our wallet address is now embedded in the withdrawal

credentials i didn't know that is that yeah i think i think it's still optional i i haven't done a um beacon chain deposit for a little while so um i haven't tried the new workflow so but i think um what you will put in there is your your ethereum withdrawal address now that will be the standard are you entering that into the deposit key generator right oh interesting yeah so that's that's totally i might i wish i might look under my counters and see if i can find 32 east today and make a quick deposit i can send you some girly if you want to do that you know i'm sorry i i can sell you some girlie i can do one-for-one exchanges one

eighth for one girly if you've got it um but no that's really interesting i i didn't uh i didn't know that that was happening that way i so but there is a flag that we're going to be able to specify in our like config to say which address we want to take payouts to isn't that right so you got two different configs here so your block rewards need to go to an address so the coin base in the in the block so when you when your validator proposes a block the um the minor tips as they used to be called i don't know what we're going to call them now but the validator tips um need to go somewhere so that uh will go to your

um an account that you can specify in real time and you can change and so that's a parameter to your validator or to your beacon node or to your e1 client you can put it in any of those places and it says when when i make a block the coinbase sends the the minor tips to this ethereum address that's me okay and then you've got withdrawals so the rest of the beacon rewards the rewards you get for for um a testing sync committees the other block rewards which are not the minor tips they all accrue on the beacon chain so you have your beacon chain validator account and number go up they are east too right exactly at some point you're going to want to

get get those out of the beacon chain and into a regular ethereum account and so if you've got a 0x01 credential no problem the beacon chain already knows what address you want to withdraw that to but if you have a 0x00 credential that doesn't relate to any eth1 account today so you need to do some kind of um conversion i'm imagining like some jim mcdonald do magic where you submit your 24 word seed phrase and it gives you an eth one address is that am i anywhere near what might work well you so you will you'll have to send a transaction on the beacon chain which will say i want to change my credential

from old style to new style and so you will supply your seed phrase that you originally registered which will give which will generate the uh your old commitment sign that and then um uh that will be validated on the beacon change so that okay that knows that you're not doing hanky-panky and then uh and then you'll give the youth one address that you want to replace it with do you remember the a ab transactions from olympic to frontier before my time dude okay so that that's what it sounds like to me um so if if you were mining on the olympic test net uh to to claim those coins on frontier mainnet you had to do what was called an a b transaction where

you kind of said yes i'm this person and i want to claim these on mainnet is that right okay yeah it sounds a little bit like that yeah so um cool yeah and there will be plenty of tools around to do this i mean clients will do this it will be like sending a voluntary exit uh message to the network now uh you'll be able to send a sent change my credentials message to the network hey so i i really like i had no idea that all of that was going on i guess i just missed it but i just remember how contentious ox01 was in the beginning and now it seems to be the default uh withdrawal style so that's interesting yeah there were some other ones um yeah it's going a bit of rabbit hole that but uh yeah it seems to have become

the the default option cool great uh yeah sorry if i got hung up on that i just i learned a ton what are we looking at here uh history pruning um yeah so this was um uh there's an eip four eight eight one lots of fours and eights in eips recently which gets very good somebody worked hard to get this number it's a vanity eip um this is a bit of a technicality uh there's a move to allow ethereum nodes beacon those and um the uh execution nodes to forget history after a while forget after a year they

can basically drop block history that there's more than a year old and the the assumption is somebody somewhere will store it block explorers will have it or it'll be on uh ipfs or you know it it it's not going to vanish it's old enough that it will stick around somewhere um and the the interesting history is the more recent history so up two-year-olds say but the problem is that the the way that beacon clients currently work is they need the history of the eth1 chain going back as far as the deployment of the deposit contract which is like um 15 like 18 months ago now so um so if we drop that history then then we've got a problem you won't be able to

sync up a beacon client so um the solution to that is to sort of uh maintain the current state of the deposit contract and serve that as part of a sort of a sink rather than have to draw back through the whole history and so um i think nimbus already does this and um or they they've come up with a way to do this uh and mark from the lighthouse team has sort of proposed making it a standard so it's good it's good news it will just make thinking a beacon no easier yeah but it is under the hood nobody needs to really worry about this i struggle a little bit to understand the incentives for that

and i i guess i've the incentives for keeping old chain data um it sounds like everyone's like oh yeah someone will keep it and i'm like okay but that's that's probably discussion for another day well it's not so useful in uh um in proof of stake land because we have this weak subjectivity thing and uh which means that it's it's possible to fake old block history um so you know once you once it's a certain age old you can't trust it anyway necessarily um so it's um there's less interest uh in it but uh yeah people you know um over time storage costs only go down um

people will store it it doesn't have to be accessible oh yeah i mean i'll store it i you know that's just what what we do but it's just uh i don't know yeah yeah it'll be up on bittorrent and you know you'll you'll be able to find some seed somewhere i mean ef might have a seed and you know whatever uh so samarasat um has just done a i i was gonna say a dirty word has done a kick butt job of uh i i'm trying to be family friendly it's my new money me um a great job of keeping up with the staking guides uh and so between him and coin cashews guides i feel really great about

uh the work that they're doing these are yeah yeah you've named my two favorites and um they're a labor of love they really are um the the amount of work that goes into these things and the pinnicky details um are is immense so uh just brilliant uh and summer has been at this for uh a very long time since before um launched a beacon chain um and you know it just kept kept on going he was doing test net guides uh yeah before launch and to know that he has open source these uh that's just it's so valuable uh here's his his topaz guide from two years ago topaz was i think the first uh test net that prism launched in you know way back in the win

so uh thanks so much to sumer that actually makes me think about um i don't spend a lot of time on it but uh eastaker is working with clr fund to um uh create a a grants round for people who are creating this kind of content this kind of brilliant beacon chain educational style decentralization content so i can't wait uh to get more out about that uh and yeah i always throw a few uh die uh his way and coin cashews way in the git coin rounds just because um they uh these super valuable service to community i mean i i used summer's guide to set up my own yeah back in the day i mean and you know i'm

taking dev i mean but it was just so good and so clear that uh that that my setup is based on what he wrote that's uh you know when people would say hey fizz can can you give me a personal walkthrough on setting up and i'm like yeah i'm going to use samir's guide and they're like well what's special about what you do i'm like nothing i'm just going to use what he does because he knows what he's doing so uh they're great guides uh evan jones is a guy who has kind of shown up lately he's um just really sharp and interesting he um has some good ideas on improving rocket pool um or or on you know promoting the use of rocket pool um yeah good good uh i did confession i

didn't read uh that article deeply i sort of skim through it do you have sort of comments or reflections um so it's hilarious i read it before vacation i remember when i when i read it it's one of those like wow this is really valuable um but at this moment i i couldn't summarize it accurately i remember he uses latex which it's so well formatted um but essentially he's he's talking about the value proposition of of why um our ethan rpl are valuable it's definitely worth a deep dive um thanks to evan for putting that up cool i'll do my homework and you took a look at rocket pool's new

terminal ui how did that go yes i love it um you know we were talking about how great samir's guide is for simplifying installation but rocket pool just has this in the bag like literally download their installer hit go and it's it's really a enter enter enter into enter start um this is as close as anything i've seen to single click installation um you know we kind of realize at this point the bulk of new deposited ether is not coming from uh full stakers it's coming from institutional investors i really think there's a lot of

opportunity for small stakers you know 16 17 eth to get into rocket pool and profit really well and to centralize the network so um thanks to rocket pool to joe clappers for setting that up yeah i think if i weren't already running my own nodes i would run a rocket pool node if i was starting today i i would do that um because i'm sort of running solo and have been since since you know the beginning and rocket pool didn't come in until a year later um i don't want to really mess with that but uh if i was starting today i'd definitely run as a a rocketable host yeah it's hilarious that you say that because i have um i have both i've two servers one of them is running rocket pull and one of them is running solo um

and i tell myself when uh when my solo node when something goes down i'm just gonna use youtube docker and switch them over to the rocket pull mode because it's so much easier to admin like it's great to know how it works and yes i know how it works now so let's just make it easy yeah that's cool i should think about that more adrian sutton do you know this guy adrian yeah he's right um adrian is a force of nature it's extraordinary um he's been pumping out some goods uh good articles lately on his sort of personal blog he goes you know he'll go a few months without something and then something will grab his interest and he'll just uh write up some uh some great little articles from

his experience of uh building teku um and this one was you know something we've known about for a while but he got the he had the opportunity to get the graphs and you know get the uh the the data on it um so we added 10 000 knows prata testnet is getting small compared to mainnet now uh as client devs we we thought it was good plan to um beef it up a bit uh you need your test net to be a real test so uh we added 10 000 more techy nodes and i guess the other teams are adding more um or when i say no i should say validators uh we're actually running them on two beacon nodes so 5000 each um and so

those deposits were processed in you know when went through and were just uh building up and adrian just measured the cpu load and other aspects of the the performance on the node as the number of deposits increases and basically until your um node is handling more than 60 or 70 validators the load increases with the number of validators that that are active so if you've got 10 validators and you add 10 more you basically double your load and you double your network bandwidth um that you're using but it tops out at about 60 to 70 and so if you got 100 validators and you add another 100 you you don't add any more load to

your your node i mean or in imperceptible extra load so you know and you can add a thousand or 5000 and it's basically no extra work for for your node uh and the reason is interesting it's because um the way we have these subnets so the validators have to listen in on different network subnets in order to aggregate attestations across them say broadcast their attestations on subnets so the beacon node that the validator is is on or attached to has to subscribe to subnets so when you've got one validator you're basically subscribed to uh one subnet and if you've got two you subscribe to two but there are only 64

of these things and once you subscribe to all 64 then you don't need to subscribe to any more and most of the load is validating gossip on these subnets so it's just checking signatures of um gossip messages on subnets that's most of what your beacon node is doing most of the time yeah i i could see someone who who wanted to break the narrative kind of say uh you know clearly it's better if you run thousands of of nodes on a validator or on a node thousands of validators on a node um you know there is i guess some point after 64 that it becomes cost effective to run more validators but i'm not sure that really disrupts our narrative too badly yeah it's it's a sort of slight

fly in the ointment of the design i would like it to be you know the old one cpu one vote you know one one unit of cpu work one validator uh that scaled perfectly would be the perfect design but it's um hard to to achieve and um the cost of running a validator and the the cost of the stake um dwarf any savings through aggregating you know yeah that's really where i wanted to go like you're still you're still making substantial profit by one running one validator on one node at home yeah um don't let don't let that effect size make you feel like somehow you're you're being wronged because it is it is extremely valuable

yeah yeah and so there is a it's a tiny increase in centralizing pressure you know the costs of of a staking operation running uh thousands of validators per validated cost their per validated cost will be slightly less than my per validator class but it is fractional is dollars per month it's not thousands of dollars per month right yeah uh staking on ethereum 20 april 2022 uh not his time not a ton has changed um i still promote solo staking above all else um i highly encourage rocket pull mini pool operation because that does decentralize the network i do encourage buying our eath if you are stay if if you have you know one or

two eat the steak and i do discourage centralization and uh staking with exchanges not a lot of that has changed the only thing i i that is sort of beginning to emerge i think uh is uh we need to begin looking more at uh execution layer consensus assistance execution layer diversity uh and i'm sort of encouraging people to try basic now yeah yeah that's a good call it's going to be the next big battle but basically is getting there yeah they've got their um the bonsai tree thing which is self-pruning so you don't have to take it down to prune the database that is fully supported and um and released

they are working on snap sync which um will is the final bit that will make it a you know fully fledged mainnet client in my view yeah and that should be out soon people ought to realize that that we all believe competition is healthy healthy in this space um that that geth has a great market share that's fantastic choosing a different client isn't slapping death in the face it is stimulating everyone to uh keep producing at a high level so yeah we're excited the big man anything he writes i'm just obsessed with yeah it's terrific isn't it um some good good stuff i mean fairly straightforward

um explainer overview of of you know where where rewards are coming from post merch so um nothing uh particularly surprising or uh to rock the boat there but really nice if you're you'd like to see how it all works uh so yeah let me tell you the thing i love about jim's writing is uh as i'm reading it i'm like oh yeah i knew that oh yeah i knew that and then i stopped reading it and i can't explain any of it like so so it's it's written so well that i can interpret it you know really well and then i i'm away from it and i'm like wow that's too big for my head uh but i i appreciate jim so much it's a gift it is a gift verkle trees what do you got here

yeah so this is um early stage stuff so part of it is um the plan the future plan is to move to some degree of statelessness in ethereum so that uh ethereum doesn't have to deal with this massive um state all the time um or not i don't necessarily have to have it all on my node at home you know miners might or i should say validators might need it and part of that is moving to this vertical tree structure so there's uh ethereum nodes have what's called the local patricia tree what we're trying to do here is reduce the witness size so the idea is that when you create a transaction you not only send the transaction you send the

data that that transaction needs to execute along with it so you know if i want to send you a uh one eth then i need to send my account balance i need to send your account balance and i need to send proofs that those are correct oh yeah okay uh account balances because you'll know my node or your the nodes out there may not know that information is part of the state they may not have the state so i've not only got to prove that i've got enough e to send the transaction but and you know so it can calculate the the the balances afterwards um and it's the proofs that are the problem with merkel patricia trees proving my account balance is is this massive blob of data this merkle proof

and proving your account balance similarly so um the vertical tree structure is a much more efficient compressed um proof that the data that i'm supplying is actually correct um and so that um the ethereum can take it you know not on trust trustlessly take that data that i've provided verify it and execute the transaction even though it doesn't have the state it knows that the state that i provided is the correct state because the proof is correct and proofs with vocal trees are very small so i i'm going to openly confess i'm not ashamed of this it's i'm i'm a normal human when we get into merkel virkle and

patricia my eyes just glaze over and i'm like i don't know those people at all like it it's um it's good stuff i'm sure a little over my hair yeah i'm trying to write a book on this stuff so yeah and i'm i'm trying to read it but all right oh christine kim i loved her tweet uh from the other day yeah christine is doing great i'm so um so happy with what uh christine's up to so we she and i did a podcast for a little while at um uh coindesk when she was there uh which was great fun and she is definitely one of the um uh one of the you know the good people

to have around she has a deep understanding she really follows the ethereum space um carefully and critically i mean you know not she's not a maxi but she evaluates claims and you know is a genuinely honest and um well-intentioned reporter on on on the space so yeah i've been very impressed with what christine's been uh working on uh since she left uh queen desk uh and joined galaxy digital and she's done a couple of really good research reports yeah i don't say this too often but um she she is someone that i would really encourage everyone to drop and follow on twitter right now because her the quality of her content is just

uh like i read this and i'm like damn i wish that i could i could put that out like it's it's really good stuff um she's christine underscore d kim uh on twitter and i lovely pull it up real fast and this isn't a promotion i don't know her i just uh started following her a few weeks ago but uh her stuff is just really high quality um i do follow her this is an anonymous account or anonymous whatever but uh yeah anyway uh just worth checking out yep seconded misconceptions you know as people on the who do this every day i the misconceptions like they're the same ones and they do get old but it really is important that we forget that we are not the intended audience

that there are thousands of people out there who really need to know the misconceptions yeah like the gas prices will not go down after the merge um that's probably the one that is is most valuable to understand yeah and the extraordinary thing about this which i didn't call out is that the first article the three misconceptions is in fortune magazine i mean not your regular crypto publication right um this is this i'm seeing this stuff more and more broadly across the media in general and um yeah we're becoming main mainstream it's rare that a day goes by when even on the bbc news over here there isn't some kind of crypto story and yeah as we said usually somewhat negative but you know it's

there it's present it's it's in the media as is in the um the zeitgeist uh so yeah we we're going mainstream big contrast to like four or five years ago when i was looking at making it my full-time job it seemed very niche yeah definitely um yeah my last job i i was literally emailing people private keys and i'm like hey uh if you can claim this you can have it and people are like please just just focus on business but yeah things things are changing quickly i love uh i love vitalik stuff this was is probably one of my favorite things he's ever written um and it's essentially

saying we can't predict the future so we've done our best and we're pretty happy with where we're at yeah yeah it's been an incredible journey i mean the ethereum way of doing things has been to explore i mean we have genuinely been explorers in a in a new found land um just you know finding our way and yeah being plenty of dead ends and you know roots that we probably could have taken a shortcut on all of this but that's how it's felt yeah that's if if you look at uh the way the merge will be implemented if we if we knew five years ago what the merge was going to look like

today it could have probably been developed in three months it's all of that that time moving toward it the research the exploration the understanding that that is uh it was necessary uh and now you know with building on shoulders of giants it's available and it's it's you know it'll keep on improving yeah yeah vitalik's a great writer i love it yeah i i love it when i when i can interpret it which uh simplified sync secrets i'm struggling with that myself oh good because i you know i have this thing where i'm like okay i'm gonna do this one i'm gonna i'm gonna tough it out and

then i get to like the third sentence and i'm like i'm out it's not happening sorry i tried yep uh do you want to tell us about that real well it's one of these problems that we've been working on for like uh three four years um there was a call put out by the ef i think it was like four years ago just to say has anyone got any ideas how to solve this there was a bounty on it um because nobody had any idea so and the issue we want to solve is that if i'm the proposer of a blog i get some notice that i'm the proposer of a block but nobody else on the network knows that it's my turn until i propose a block and then everybody can ver verify that i was yep

it was ben who was uh supposed to propose in this slot and what this avoids is the idea so it feels if i know when you are due to propose a blog i can denial of service your node um potentially and prevent you from proposing that block uh and in in that way selectively you know if i know the next 32 block proposals i can selectively dos them and then take control or exert control on the network so the the idea is we want this secrecy around who is the next block proposer they know but nobody else does but then you can verify it and it turns out to be a really hard problem and um some of my colleagues in consensus came up with the very first idea how to do

this using some zero knowledge proofs but um it was sort of heavy weight and over time it's just iterated and iterated until we've yeah vitalik's come up with something which yeah maybe it's it could work i mean it's hard to say just now but uh it's been really interesting to see progress on this that's great uh yeah so i'll take that summary implementers call 85 which i haven't heard go ahead well it's fun to see christine uh is taking notes on this as she does for the

alcohol devs cools as well um this is christine kim we mentioned uh earlier um and it's fascinating looking at her nose and my nose and they might be of different meetings but we just have such we have different viewpoints on the same meeting and we re different things pique our interest this is actually what i was trying to reference earlier yeah so i um that's great though like it's great seeing the different yeah i love how uh you both have different takeaways but as you know as a third party reading both of them i find interesting points from from both points of view like everything doesn't interest me but i'm able to pick and choose what i found most interesting i still want to go back and catch that one all core devs

christine kim's notes again yeah so that's those are going to become very valuable obviously tim does uh after every all-course devs call he does a tweet storm that uh is just uh if you if you ever miss an awkward even if you even if you're there the uh his his tweets kind of put it together and and uh they're super valuable they're a great reference um dev connect dev connects yay uh it's like next week um wow gonna be my first outing for a very long time yeah uh you will not be there is that right i'm skipping we uh we weren't ready for me to travel to amsterdam uh so i'm staying stateside

uh yeah well shame shame i look forward to seeing many people many friends uh there but um i've already told my family look there's it's a certainty i'm gonna catch kovid right so just get ready to isolate me when i when i come home because um it's gonna happen but have you had it yet no not not knowingly so i've had a bunch of vaccinations so uh hopefully it will be um be all good but uh yeah um we're essentially the same like i've had all my vaccinations we've pretty much stayed in the house and my wife's like you're you're not gonna throw away three years of isolation on catching covet now i'm like okay i think the the time has come the time

yeah i agree i want to for anyone who doesn't know uh east staker and you can go learn more easttaker.cc is hosting a huge uh i don't know what i call it huge a valuable um staking three day kind of many conference there as a part of devconnect they i think i know that they started with 360 open seats uh they are asking for a hundred dollar retainer for the tickets and then you're gonna get that back when you actually show up you may not know that people like claim these tickets and then don't use them uh so the the speakers list uh the contributors there are it's just it's an incredible kind of lineup it's

it's really the thing that um makes me wish that i could go yup yeah i think adrian sutton we mentioned earlier we'll be presenting a little bit uh on that um yeah i'm very much hoping to get to the east acre event um the problem with devconnect is there's so much going on um so the first two days we've got like um protocol dev teams l1 and l2 get-togethers which are sort of mandatory bits for me um and then day three i'm at the light client summit day four is thursday um yeah bunches are hoping to get to east daycare it's the hackathon that day right so hoping to show up at the eatstroker

event on the thursday at some point yeah that's that actually kind of one of the things they're going to promote is that clr funds uh raising uh grants round that's going to begin right after devconnect and one of the things that we're hoping to do is get some interest in that hackathon and then pivot those projects into uh clr grants matching i i think it's a tremendous opportunity the thing i'm most excited about seeing the outcome of i have no idea how this is gonna unfold but they're doing a um explain ethereum like i'm a preschooler explain it like i'm a five-year-old explain it like i'm

a year old explain it like i'm an undergrad explain it like i'm a graduate student and they're they have different speakers for each of those levels i i am so hopeful that that gets put together into like one quality video distributed because it's excellent it's a fantastic idea brilliant excellent yeah i think uh jim mcdonald is uh we'll be speaking as well um yeah yeah correctly yeah so so many like uh and i guess so i'm i'm not going to be there obviously but i have kind of watched from the sidelines as uh the the other eight each stakers have just busted tail on this for the past four months so i'm really confident that it's going to be a quality project so yep i need to make sure i show up at

some point to get my hundred dollars back but there ought to be a poep too that's really the why why go if it's not it's funny like i was like um honey you know there's gonna be a poep and i i really need to get it and she's like well can't you just buy it somewhere that's not the point outrageous yeah well there we are we've done we've done an hour um who'd believe it oh dude we've rambled on for an hour i if i were to guess i would have guessed that we had run long but um i'm really glad to go through this stuff with you it's it's fun and uh yeah thanks it's so it's great fun love talking through with you yeah thanks everybody for

participating i'm gonna end the recording somehow