What's new in Eth2 - 11 February 2022 with Ben Edgington and superphiz

Transcript

hey everyone i'm superfiz and today is what is today february 14th it's valentine's day uh 2022 and i'm here with my good friend ben edgington hey happy valentine's day dude yeah my uh so my my goal for the day is to not be in front of screens and to spend time with family and um i guess you're my family today ben uh yeah indeed and so let's uh keep it tight and we can all go and be with our loved ones awesome yes um so i i felt very good about what we did last week um and so i think we'll just zip through it but before we begin i have to acknowledge that um

this this naming i'm gonna let's see i'm gonna move my s i like to see you when i chat with you there we go um oh that doesn't work sorry i'm getting off topic here okay so the naming convention thing uh is something that pops up as a matter of fact when i when i posted what's new in eth2 on e-sticker read it this week um there was like backlash like that's not correct and i'm you know my my response is ben knows what he's doing uh he has a wealth of information and i love learning from him so if he wants to call it eth2 if he wants to call it e37 i don't care because i'm evaluating your whole corpus of information and if it's

valuable to me then i accept it and it doesn't matter so much to me if if you choose a different naming convention but for for everyone's benefit can you talk to us a little more about about this uh this stand that you've taken all right let's do it um yeah and be nice to me please because i've just come out of two hour workshop with for with the tank rad and vitalik on dank sharding and uh my head is spinning yeah i i missed that but how was that about let's do that next i guess yeah it was heavy but we'll come to that uh on naming so i still i i have this ethereum two thing in my mind um

uh and ethereum two has been a thing since um 2014. uh vitalik first called it that it was to be a big cryptographic um upgrades to ethereum um you know proof of stake came in under that banner lots of scalability and stuff eth2 has never been a specific deliverable there's never been a thing we've said this is ethereum too we're going to deliver it it the the concept's always been a big upgrade to ethereum what are we doing we're doing a big upgrade to ethereum so i i call it ethereum two um the ethereum foundation has come in and said we don't do that anymore and that really annoys me right i mean so my main objection number one on the list is uh just kind of a high-handed way this has

been done right the ethereum foundation has just come along they said oh devs don't call it that anymore patently not true um and um and we're deprecating the naming well i mean you know that's not how things work in this ecosystem um no no consultation was done no community engagement nothing it was just like we're not doing that anymore so i immediately immediately riles me i you know i'm gonna stand up for permissionlessness in this in this context um of all things so uh as the purveyor of what's new in eth2 you know you're gonna have to do better than that to change my mind um and i want to share a contrast to that is my observation is it has been a very slow and deliberate shift toward

the consensus and execution layer language um i didn't feel as if it were imposed upon me but i do see i do see uh eth2 as being a a large set of global upgrades not just the merge and so many people equate eth2 to the merge um i can see how that can be confusing right right and you know it's not a clearly formed concept in my mind if you ask me define what e2 is i can't tell you it's x y and z it's a big future upgrade to to ethereum um and um and just to be clear i don't object at all to the consensus layer and the execution layer naming except that kind of cumbersome but you know from a technical

point of view it's a nice distinction um and it's um more precise and heath one and eth2 clients though i always go back to that in my own mind because somebody says execution layer i have to do a few micro milliseconds of processing to oh yeah he's talking talking about the ethon client or whatever um but you know that's a technical accuracy issue uh and i i don't really have a problem with that they're not you know pretty terms but that's okay but as a um a general concept of what we're doing i i like ethereum 2 we it deserves a big upgrade naming because we are doing the largest upgrade that has ever been

done on ethereum um and uh i i think it is different you know ethereum on proof of stake ethereum with all the innovations that we're bringing is different from ethereum that we've had so so that's my thing um yeah and uh i'll i'll uh i'll keep on plugging away and if people can i pride you on something so um i i like the i was just testing the screen share but i do want to uh this is faster it will leave it for now um i get eth2 as the future of ethereum and i would actually be a little i would be cool if we said um future projects are east too uh so that you know sharding sharding could be

east too that that fits my question is after the merge would you then refer to the merge chain as an eth1 chain and an eth2 chain or would you then refer to the merge as ethereum and future projects as youth 2. well yeah that's a great question i mean and i don't have a clear answer i'd say we have we we're in the eth2 era the ethereum two era at that point and there's a whole bunch of deliverables that kind of go into that and then at some stage we'll talk about ethereum three and uh and we'll sort of move up move up the the versioning um tree and you know some people have uh yeah we we have this happy banter on

twitter and you know proto-lambda sort of um has made a um what's new in consensus layer yeah um a site which kind of mirrors mirrors this and that i bridle a little at that because i'm not i'm not i'm not writing about the consensus layer that's not my constraint my vision is is the whole thing it's the scalability it's the um the the proof of stake it's the the whole enabling of the future of ethereum and it always has been it just happens that the proof of stake is the thing that we're delivering first and that's kind of dominated in the narrative but uh if you look back through the history of what's new and he's do it's always been about more than the beacon chain and the upgrade

it's um it's about the the the big picture okay so that that was going to be my last question but i have one more if if you view what's new in eth2 as and i i hope that you understand my questions don't tend to poke you i i am deeply aligned with you and we we click really well this is truly about understanding um if if what's new in eth2 is what you would consider a kind of global ethereum newsletter then how would you it's certainly a different newsletter and i don't consider it a competing newsletter to evan and andrew's newsletter but

it's certainly more future oriented is that are you more focused on the future where do you feel like they're more focused on the current i'm focused on things that interest me which is much more limited set of things and goes into weakened ethereum basically protocol so um the protocol innovations and the future of where the ethereum um layer one protocol as it were is going i really you know don't have the bandwidth or capacity to talk about roll ups and you know application level stuff and all of that um i i should just just say a quick word on on the branding i mean so you know what's new and heath 2 um so an interesting experiment was inadvertently done a while back when uh

what uh weakened ethereum uh cheekily rebranded what's new in consensus layer and my readership dropped by half right because the name recognition out there is for ethereum two i would much rather p reach people with good information under a slightly wonky brand than not reach them at all um and there's a danger if we just say oh that this is ethereum and then people you know if there isn't a brand around it then people are not finding good information uh out there um and so i i feel that by deprecating the ethereum to naming we're really missing a trick it's always been filled with different content uh for the last eight

years it's had different content every every few months um but now just so we're not doing ethereum two anymore i think is missing a trick here from a branding marketing um reaching people communication more importantly point of view my honest perception is that um people are discounting the content when they see eth2 because um of the hot button issue and they automatically assume that you are on a fringe obviously i cl i follow you closely and i recognize you're not on the fringe but i i do i do sense that people uh see

that as a dated kind of um fringe topic that they they don't align with like they might have in the past so that's that's just food for thought interesting i support you no matter what you put out that's not true actually i'm picky but i'm just okay with that one i and you know i don't know how long i'm going to be writing this thing for um it's been three and a half years getting on for now and um uh you know at some point it will just come to a natural end maybe after the merge i'll just like shelve it and say enough um and then that will save the whole rebranding thing yeah i i hope you don't end it um i

don't care what you call it but it is it is definitely one of those things that i drop everything for every week and it's not that i i investigate everything deeply but i feel like at a high level it helps me i i miss things and i'm sure you do too like like i didn't know that happened um and so it's one of those valuable catch-ups so yeah anyway that's a chat people wanted to hear i was curious about it and i'm sure it will come up again but um yeah thanks for talking about it um let's see so i did some highlighting here uh people asked me to share the screen and i thought that was valuable so as we go

through i will be opening some links and just sort of uh trying to follow along visually for people who are watching my assumption is most people are going to listen to this while they're driving or whatever but you never know um let's see we did this last week so yeah i put out a call this is actually based on a conversation that i had with some folks from what is now coinbase cloud which i used to know as bison trails where they said we need data-driven analysis our our clients need data-driven analysis and so if you give us that data we will put it in front of our clients and help them make more informed decisions

about client diversity um and so i hope people still take me up on that call i've actually begun issuing a co-app that i think is gorgeous uh can i i don't know if i'm going to be able to find it uh yes i need to find it and i would go to hold on i don't remember the damn name of it okay anyway i'll find it later um but i am issuing a poet now ben has oh that's how we find it we go to a uh sorry i'm just so proud of the design

uh i've been trying to pop to people who make significant contributions to eth2 client diversity and i really want uh anyone who is so if you've made a significant contribution reach out um let people know that hey i think i should have gotten that poe up i would love to get this to anyone who has made significant contributions and there are there are a bunch of people um i love the design of this poem it kind of speaks to the research angle of contributing to client diversity which i think is so valuable that is a topo app that is great and i am a proud possessor of that part so thank you very much

yes i'm so glad to share it with you um i mean i i did find found the client so you know i think that's that's always an interesting thing you know it is i i recognize that you're in that unusual spot where you're on one hand a uh client owner and on the other hand you're also supporting this thing that that is good for the health of the network um i in in my honest i i would like to believe that if teku were the majority client the super majority client that we would have very very similar discussions because this isn't about winning it's about true diversity absolutely are you um

uh happy with the responses you're getting to your client diversity um uh the the the paper writing uh initiative well so joseph cook wrote the first one and so far it's the only submission and i'm literally obsessed with it i i didn't know joseph cook before this um i know he's a uh court dev apprentice um but the quality of his paper and it's linked here i hope everyone checks it out um the quality of his paper is just so do you know what it is it's it speaks just to my level like i can read it as as a as a normal ethereum user without feeling overwhelmed i understood

everything in this there weren't topics where i'm like ah that's a little above my head i'm gonna skip over it um so i felt like it was just so well written and um i do hope to get more submissions one of the other not a paper but another submission i got and i think we'll talk about this eventually too is uh invisible symbols uh website pools.cloud but we might catch that later in the in the chat yeah that i um saw at the weekend it didn't make it in time for this edition i think but this is good yeah sure and we'll catch up on that but yes uh i am i am satisfied with if even if that's the only submission i'm satisfied i'm content i would love to see more i i am all about building on

success um so yeah um this merge readiness checklist i know you're gonna be you've been involved with this some some level what are we looking for here yeah so i think the uh really cool thing about this is that at the top of the list which is this kind of work that needs to be done the prep work you know the client development the spec writing and the kind of spade work on that basically all done um there are a few empty check boxes there but they are either nice to haves or not particularly difficult there's no blocker whatsoever in terms of the technology or the client implementations to to doing the merge or

all of the empty boxes are uh down the the bottom half and they're all testing so it's just testing testing testing um and that that that's it between now and merge as we've got left which is uh astonishing absolutely astonishing place to be it really is i you know i still i think about this the thing that stuck in my mind most was the slog for figuring out uh ssd uh because i listened to those talks for like weeks and i'm like i don't know what's going on here um and to look back at that and then compare it with where we are now it's just such a wild contrast like it's we're almost there um yeah bike shading about serialization for for six months yeah i really i i've revisited that for

my uh my book chapter on ssc which is mentioned a bit further down recently so if you want to relive the saga you can you can read that uh maybe i'll understand it the second time around so kitsugi and we know kenzuki is sunsetting it's not too late to join um we've actually i've been in some chats with uh an interest in getting a co-app for that because i feel like it's it's the first public test net for um the merge and it's kind of a neat thing um but as at sunsets i this is the biggest takeaway from me um don't spec your hardware based on the happy flow be prepared for increased resource usage during times of non-finality

yeah yeah we've seen that before as well on on test nets with just beacon nodes and uh when the chain is not finalizing when it's under stress uh typically there will be many several forks and clients have to maintain data on all of the forks they can't sort of finalize it then prune it all away that that blows up your memory usage it blows up your disk usage it blows up your cpu usage um hopefully you know we we will never see this we've never seen it on the live beacon chain we've only ever seen non-finalization on test nets but if we do end up in a situation of non-finality due to a client bug or

something then if you're on minimal spec hardware you're really going to struggle and you'll get penalized for being offline during that period this this is the leak the quadratic leak process so you know raspberry pi's and all of that are very cool indeed it and it's amazing it's incredible that the the whole system can run on a on a raspberry pi but that's a happy flow and that and ideally you know in a perfect world that's how it will always be but just be aware that um that things do blow up when that when it gets turbulent yeah and so i i remember vitalik's original promise of the beacon chain um was

run it wasn't a talk i want to say it i'm not sure but his talk was um you know you're going to be able to run it on a laptop and you're not going to worry if you shut it down sometimes or if you start it back up because as long as you're online two-thirds of the time everything is fine that obviously matured to the point where yes you could run on laptop hardware but we really want you to use you know a dedicated machine and yes you only have to be online two-thirds of the time you're probably going to want to be online as much as you can be um but you know a lot of my friends are running raspberry pies i think of you know there's this um this clique that enjoys running on minimal hardware

do you think is it a poor choice or is it acceptable given the the breadth of the network it's a personal choice i mean if you know the risks and use and you um take that into account then that's fine you know if you're prepared to transfer your validators to a beefier um piece of kit uh if the if there's a problem in the network or or whatever if you're confident with that then you know it's no big deal and uh you know you're doing uh doing the planet a favor by running it on low-spec hardware um and and that's all good but i i don't want people who are uh naively assuming that you know this you know

this 50 bit of kit or whatever is going to be fine for the next three years or what um uh for for production just because um it it probably won't that's that's very fair um and i'm the same i don't want to discourage people from trying those things uh but i also like personally i use more powerful hardware because i i was i was involved and not involved but i was an observer of the madasha non-finality incident and i you know you had to have horsepower to grind through that but i i still agree people can be okay if if there's a non-finality issue they

will experience quadratic leakage they're not likely to be pruned and as soon as things are back online they'll begin recovering i guess we'll just have to see how that plays out i yep yeah i know it'll be running for two three more weeks uh while we get our ducks in a row for for kiln um we've got a couple of devnets to run for kiln before we make that um the the new public main net um and the reason we're replacing kinsugi is there's been a bit of a spec upgrade not backwards compatible uh the idea is that kiln is the final spec this is the sort of frozen version and it's the last test net before we start forking like uh rop stern or

wrinkly or whatever the um ethon test nets are uh and um forking them to proof of stake and fairly it's it's the the hopeful final client right because we can always always upscribe it scrubbing kiln and trying again but uh it will be a great sign of progress if the specus frozen kiln is successful it will be that a tremendous green light for moving forward right yeah totally yep yep so east sticker made a video about joining kentucky with tekko besu will do the same uh when it comes to uh kiln and as well i haven't seen this but um can you tell me about this one

yeah we did a consensus webinar on joining uh kinsugu itaku and vesu so um uh adrian sutton is is there as is gary schulte and uh um and i pop up to do some q a so i mean just a word on baysu it's not a well-known ethon client execution client um and and to be fair has been focused more on the enterprise side of stuff on the private network sort of things for um its first couple of years of its life it we are now focusing it very heavily on mainnet support we've got uh incredible team around it lots of resources really um making it uh a

first-class mainnet client so there are new features like bonsai trees which uh reduce the disk footprint and it's auto pruning it's by its very nature that by design it's self-pruning so you don't have this thing like with geth where you have to take it offline every few months to prune the database uh if you're running out of disk space um with basically it it automatically prunes itself as it goes um and uh yeah uh just um putting in snap sync and all the um uh final sort of ether one features but in the the client diversity conversation goes beyond your consensus layer client um you know it's important it's equally

critical we have that diversity in the execution layer clients um geth is the one that everybody reaches for naturally and i fully understand why but here is an opportunity to um to to diversify um on you know never mind or um or base or one of the others just to clarify my understanding basu is is now a consensus product is that right uh always has been so uh i was uh i was in the very first sprint planning meetings back in uh early 2018 i think we started maybe late 2017 we started the basu project it was called pantheon for a while i mean when we started it was called new client it's always been mainnet compatible but

um the focus of engineering resource has been more i think it's fair to say on the uh private network side we've always maintained mainnet compatibility but now the emphasis has shifted right back i mean for the last six to nine months or maybe longer than that actually tim baker was product owner of basically for a while so if you need some pedigree you know he ran it for a year or so um and under him and and successors um it's being very huge emphasis being put on uh excellent first-class maintenance support it took me a long time to really integrate basic into my into my awareness because i it was it was hyper ledger basically for the longest time right and uh yeah still is um it seemed foreign to me i guess for some

reason um and i i think now i'm i'm coming around to be more willing to um to promote and speaking of reddit comments there's a reddit comment this week like why is there so much talk about layer diversity and not as much talk about execution later diversity and i'm like that's probably because some of us have been screaming at the top of our lungs about consensus layer diversity but it doesn't make execution layer diversity any less important so for my part anyway i feel like i can focus on one big goal if i diverge and focus on both goals i'm not sure that i'll ever

be a part of achieving both but i do recognize the importance of um execution layer diversity as well yeah well that battle to fight when it comes yeah it's gonna be an interesting one speaking of hardware spec um like raspberry pi um i love what ethereum on arm is doing i i imagine you're also a fan they they are incredible i mean the the engineering brilliance there is is superb um we have done a couple of things to make sure that teku is kind of um compatible in terms of our dependencies and so forth but they've really done all the hard work on on their side and uh um the the stuff they're able to to pack

into those uh little uh things is incredible um yep do we do we know who that is or who they're associated with like i recognize them from twitter but i don't it doesn't cross my mind who it is uh don't know um yeah that's fine and we've chatted a bit about kiln um kiln spec one is is out we have another progress tracker for kiln right following following the emoji format yeah so we have the fire brick and yeah people are making progress we can't go beyond m1 so m is a milestone so we've got like six milestones uh there i can't remember exactly what they

all correspond to but milestone two is implementing the v2 spec and the v2 spec isn't out yet so we uh i mean we basically know what's in it but uh um the the detail isn't final so everyone's just paused on milestone one for now and then yeah by the time we get to milestone six that's the full gamut of test cases done i guess when the v2 of the spec is released this could actually move fairly quickly uh is that a fair assumption yeah yeah uh does why not very exciting um i see stuff about the engine api i have questions about that but we'll probably catch that some other

times later or whatever kill test net uh looking for kiln to open up to everyone in my intention is never to ask about timelines but my guess would that would be early march um that seems fair to everyone yeah yeah that sounds about right i mean yeah so client teams need to um get the configurations in and make it easy for people to to configure um that configuration is so valuable because as end users um you know i'm not

a developer i can use linux i can i understand the clients well but um i still look for easy setup easy configuration um and that's that's what's it's easy to share with you know the masses and so the more refined that is the more people we could involve in testing yeah it's it's much easier if you've got a minus minus kinzugi or minus minus uh kiln flag on it so we try to do that just to just to make it uh take take one barrier out of the way and it's hard enough as it is uh so the next thing is beyond the merge my my real comment wait did you have thoughts here yeah i i saw so in terms of sections

that i put in i sort of think of random titles just as stuff comes in so i collect all the things that are incoming during the the couple of weeks and then i sort of try and arrange them into loosely related categories and then stick them under under a title so this is a kind of random collection of stuff which um is you know coming up um both on the uh execution side and on the consensus side i just wanted to note that we're starting to look towards what they're calling on the ethon side shanghai or that may change the name so that's the fork or upgrade if you prefer due um towards the end of this year so maybe six months post-merge um

and uh all timings to be treated with uh suitable skepticism um uh and we will likely do a simultaneous beacon chain upgrade at that point if only to enable withdrawals i mean you know as a bottom line that's the sort of um table stakes for that uh upgrade but there are a couple of other things we might look at on the consensus side um which uh yeah i just the italics blob transactions thing i've just been listening to an hour on that and um um yeah i think i need to go back and watch it again do you do that

maybe it's just me it's kind of a step towards dank sharding which is another topic um so tank sharding is kind of fully sharded data availability vitalik's blob transactions are a step that is iterative it builds on our current infrastructure without introducing a huge new um you know networking infrastructure for example um and uh gives us the benefit of some data availability boost for um roll-ups basically uh with so it's a sort of you know halfway house it's a step in the right direction that gives us some very quick benefits without too much

disruption well we while we work on the sort of the full dank sharding thing in the in the longer term we spend a lot of our time in the weeds but i think the message that people who don't get the big picture need to take away is um ethereum traffic will move to layer two all of all of this preparation all of this organization dank charting uh blob transactions are are being set up so that layer two can handle immense capacity and settle on uh the main chain and i'm not sure that's completely evident to everyone but it is something that that's that's what all of this discussion is about um yeah yeah yep so we've said from

early days that ethereum will be the settlement layer for the planet uh this this is what it looks like people it's chains that sit on ethereum and ethereum settles it's the high court um but the the the chains that sit on it the roll-ups as we've yeah the vision has come into focus you know three four five years ago we we knew it would look like this but we didn't quite know what the technology would be in detail now we know it looks like roll-ups um go roll-ups optimistic roll-ups that's where all the user activity will be and the ethereum is you know the final arbiter and people will not be transacting individuals will not be transacting on

the ethereum base layer in in years to come unless they are whales because it's so hard for me to hear that i'm not a whale but i've been transacting on main chain for so long like i can't imagine a future where i wouldn't be transacting on main chain and i know that's uh sorry having two year old there are three old issues um i know that is the future but it seems so i'm like i'm gonna do it on name main chain like that's just what i know yeah yeah i know i feel exactly the same way i haven't chosen my layer two yet i mean all my assets pretty much are still on layer one at some point you know before the gas prices um prevent it needs to move to a layer two

uh we all will need to do that but uh the kind of um the candidate the one that i want to you know sit on hasn't emerged for me yet so i i don't know i i have uh as an exercise i sort of send an e to each of the new chains like i i want something there just to have my foot in the door but i ha other than gnosis or x-die which is technically it's a maybe a slightly different beast or a manic polygon those are the two solutions that i actually do things on but when it comes to optimism arbitrary i haven't really felt uh felt motivated to act to act on those

chains yet yeah and interestingly those are the two which are not roll-ups they're the the side chain solutions and i i guess they're ahead right because there wasn't any really new um technology there it's it's just speeded up older technology whereas um the roll-ups are um implementing uh you know from scratch and new technology so it takes longer yeah zk syncs looking interesting lots going on there so yeah definitely looking forward to interacting more with dk here we go i'm not 100 sure what that's going to look like but i know it's something that i'm supposed to be excited about if that makes any sense

speaking of the the call i think this is called 81 the thing that that struck me most about that consensus layer called um was the focus on post merge and that gives me the feeling that whether it's accurate or not it gives me the feeling that the merge is more or less buttoned up and that we're already ready to shift focus to what's after that how how fair is that yeah so we talked about that was the last 10 to 15 minutes and um i wasn't the very first time we talked about that but uh but yeah certainly our eyes are starting to look beyond the merge um

seriously for for perhaps the first time and uh it is a bit of a treadmill dude you never quite get there it's sort of always another thing so uh right we get this uh incredible effort for the merge but we can't kind of put our feet up we got the next thing to do then next thing in the next thing but uh keeps it uh keeps it real keeps it exciting so um yep yeah yeah and you know people like vitalik and dan crad and justin and the ethereum foundation researchers are not looking one step beyond they're looking two three four steps beyond and uh you know into uh a future that's that's kind of years long by the time it's reached implementation the research has been settled and

they're they're no longer worried about that because they are able to uh focus on the big picture and i guess as users we're sort of stuck in the middle between catching up on two-year-old research and observing the current implementation so we did chat about uh client diversity on the execution layer and it is important that we continue to encourage people to use um other execution lines other execution clients than death i i plan to spend more time with bay su that's i've been moving toward that but i just wanted to point that out one more time ethereum pools

loving this twitter yeah it's great isn't it oh yes i i don't think i've seen it this week but uh any takeaways uh it's amazing level of detail and um it's so i think we'll come on to this a bit with with rated as well transparency um i think one of the the keys to making this space work is is transparency then people can make good choices about you know where they put their stake or um how the network you know which clients they run and so on and so forth and um i'm impressed with the level of due diligence that ethereum pools has done the

uh amount of data and identification of the players and so forth um yeah just the detail is incredible yeah i share the same i i was skeptical at first because you see you know someone you you hadn't recognized you don't know who they are but they begin dropping this data and so you're only left with data you're not left with you know someone's background or credibility and just looking at it it's so such high quality data they are spotting things that i i probably would never have have known about if if it weren't for them sharing it yeah fabulous and you know i occasionally do a little bit of digging myself and you know have my own dashboards and stuff i don't spend a lot

of time um and nurturing them but the data that i'm seeing from ethereum bulbs matches what i'm seeing on my dashboard so you know i'm i'm convinced it's it's credible it's accurate and um and reliable so um absolutely great yeah let's now is when i probably want to pump this uh pools.cloud because it is it's the same kind of network monitoring who's doing what and i i know that it didn't make your newsletter it was probably released a little bit after that i'm feeling really optimistic about this uh this i guess data collection report style do you have any thoughts on it again transparency i mean terrific um

yeah surfacing the um the the information gathering it servicing it it's uh it's all there i mean you don't have to have secret insider knowledge to to build this stuff it's a lot of it's on chain um and or available through analysis like michael sproul's analysis which i i gather is mashed up with this this data and uh yeah this is super helpful this is really really good but you get to see um who are running the mono cultures it was most exciting to me because you know i had been tracking michael sproul's data and we knew like we've got you know prism

has a super majority we've got to get it down and so i kept pushing on home users like come on guys like hey everyone if you're running a solo validator you to switch over we got to do the fight and then i started seeing ox and visas or invisible symbols where data that he had linked proposals to operators and putting them in a room together they they both just clicked their data and now um we have a much better view of where those proposals are coming from proposals are coming from and it's not people at home people at home generally have decent diversity it is these huge

uh providers that uh maybe need a a good solid push to um to make better choices i i have some optimism uh from um professional conversations that things will change uh over the next months but uh yeah i think you made the point last time we spoke uh you know the the uh prove the pudding is in the eating when we see these numbers change that's when we can start to relax a little yeah i'm hearing the same kind of um i guess the the calming phrase that i get is you're going to see big changes around the march time frame and i just it's hard for me to

to take that in stride because that's a way of saying don't worry about it it's going to take care of itself and then we let our guard down and march passes and nothing has changed and we're that much closer to the merge the merge isn't the most important thing for client diversity but it is a big kind of stepping stone and i don't want to let our guard down until we've really solved it yeah and you know if the beacon chain stops today or has an incident or whatever then the the only people really affected are stakers and stakers know what they're getting into at least they ought to do right i mean there's a lot of value staked but it's not um uh it's not running d5 right uh whereas

after the merge if the beacon chain gets into trouble then then it's chaos i mean chaos for everybody exchanges and and d5 apps and everybody who uh relies um on the chain yeah nfts and everything goes crazy so it's um it's a much bigger deal so for now we're still sort of sandboxed but uh i was horrified when when we actually saw real life slashings on the main net um i think anker had like a 175 validators slash like very early on i think that was stay staked steak. yes that's right that's right i apologize so they both had uh fairly large slashings

and i was horrified and i but i'm now i'm happy those have happened because it's sort of given everyone the realization that slashing is real like it's going to happen it's not it's not like a an invisible stick it's a fake stick it's really going to happen um and so no one really gets to come back and be angry when it happens to them because it uh it was never held it was never a secret right yeah yeah i think that was um uh i mean they made they they made good the stakers so um uh which was the right thing to do but um absolutely it was a wake up call i didn't check out raided oh that link that link is broken uh it worked let's

check it 20 minutes before i published uh it was live um but a footnote down the bottom uh there is a twitter thread um there which which goes through everything um i was hoping that link would spring to life again i don't know what happened to it so so yeah so this is um from uh uh elias who i and i'm not sure if he's still with bison trails or coinbase cloud as it is now but uh used to be but uh and one of his colleagues who uh yeah there we go yeah um who have made a dashboard um so i linked through to the dashboard or it might be

linked uh here separately so if you scroll back up again uh to there if you go to the front end there i do like doing this with screen share better by the way yes so what they've got here is the deposit address of on the left-hand column um of validators uh and then um some stats about the performance um based on heuristics around the attestation performance and you can click on the deposit address and see sometimes they share so you can see that's kraken you can see

the performance of all the kraken validators under that one and you can compare between the different pools that have deposited from the same address this makes me a little curious did i close oh okay so this says kraken has 30 649 and this one says 30 oh wow that is uh so it's good when those those correlate so i'm happy about that and you know it does depend on deposit address and you know not all pools use one address and and so on and so on but uh um yeah ui needs a little bit of work uh you know it'd be nice to be able to sort or filter um i couldn't get search working and things but uh as a concept as a you know a first step this is

another dimension of transparency you can compare the actual performance of um relative performance of the the various pools and and you can take your own validators uh fleet and compare it with uh with the others as well to see how you're doing my favorite thing about tools like this is um many end users may not see a purpose for this but others will see this and think you know i could do something like this only change this and that and then they this has become either a building block or a stepping stone to something that other people can use um and as you mentioned the ui improvements here could really if if this showed who the operator was

it could really benefit yeah yeah and um elias and uh and friends are very keen to make this a community effort i mean they've kind of done the first step of of building the initial database but they want um the vision is to gather data um from um stakers you know everywhere to add to the um you know the base of data available here and make it much uh richer and more useful well i love it so yeah i chatted a little bit about uh dvt that's obal and ssv uh versus rocket pull and i know that reddit i don't know if it's just reddit but they have this

idea that those are competing products and i just don't see i don't see the dvt stake and trustless staking polls or i guess that would be the right way to describe them as as competitors uh agreed uh one is infrastructure i mean the distributed validated stuff's infrastructure and you can build many many different protocols on it and i know i saw down here digi is going to talk about that at east denver um uh-huh there are a couple of talks about the distributed validators i think can i make a confession i haven't i haven't read it yet does that make me a terrible friend uh well it's there isn't too much there yet so the links in blue are the bits i've

done and the links in black are the bits i've got to do which vastly outweigh the bits i've done but uh yeah yeah just plowing through chugging away writing the book um what are your long-term plans for well as if you go up to the preface so on the left if you go preface rather so plan is as follows so the bullet points there you mean i could read the book to find out the answer the question i'm asking indeed um so basically i want i i what i want to do is have frozen editions which reflect at a point in

time the the reality that then prevailed so i'm currently writing alter this is the outer edition um and the spec is the outer spec and so on um and i'm writing it post facto we've delivered out i'm not writing about things that are going to happen it's too much of a moving target i'd have to re-edit it every day so uh when the merge happens then altair version will be frozen uh that will be the outer edition and then i'll start work on the merge edition which hopefully won't be a massive lift it will yeah be a few more chapters and a bit of re-editing and whatever but uh at some point i'll catch up um now i'm not going to get the whole of altair written up by the merge i've got to i've got to um narrow the scope somewhat but

yeah i'm going to have a good stab at getting uh most of part two in part well part three is done so basically most part two all of part two done by alter may i should invest more time in reading this like these summaries help me you know i have a swiss cheese understanding and these summaries help me kind of fill in those holes yeah i i mean i'm writing the kind of book that i'd like to read and it may not be maybe i'm the only person in the world who'd want to read it but i i've had some nice feedback and i found it very useful to be able to refer people to it in conversation and say hey actually i've written something about this you know um yeah i i would love to see it eventually as an nft um trent did that um

uh the nft book i found that really interesting i'm sure it was a ton of work but it was really yeah that's not really my world i'll need some help help with that but uh yeah let's talk about it uh and i don't know at heart engineer but uh just somebody who popped up on on twitter very helpfully just uh gave me a style sheet which um made it usable on mobile i i i learned all my css in 2003. um i'm not not even joking um pre-mobile so i've i'm clueless about uh app media queries um there we go yeah so research um this is balancing attack lmd edition um do you want to summarize because i

like i get lmd ghost but i uh i couldn't summarize yeah so the this is sort of a long line another in a sort of long series of articles from the same researchers have been looking at the consensus protocol we use um in eth2 which is a combination of lmd ghost and casper ffg is kind of a bit frankenstein you know it's bolted together um we sometimes call gaspar and they have been looking at different ways to attack it and you know their very first attack they came out with was sort of so infeasible it was not really um worrying at all but over a few

um iterations they they refined it to become much more effective so we responded in protocol by introducing something called proposal weight boosting um and they've come back with a new attack which even with proposal weight boosting means that an attacker who's got some power um say 20 percent of validators and some quite strong power over the network so they can organize things so half the network sees one thing first and half the network sees another thing first they can partition the network so it never finalizes you know you're just swapping between two two parallel chains uh you've balanced the the the votes in lmd and you can keep that up

indefinitely if you maintain your power over the network and and i you know it's this is again back to the theoretical but very infeasible to pull off in practice attacks um and it turns out there's a really super simple um way to defeat it in protocol which we should have done already i mean um and i noted here you just have to stop uh if if you see two different votes from a validator you just ignore both of them rather than counting the first one so um it is easy to defeat it does kind of make me wonder what's next you know it's sort of they keep coming back you you you whack one of these uh and uh i think you fixed

it and then they come back with another so i'm super glad they're looking at this um it's always it always brings me back to decentralization like i i know that that is it's my obsession but with with a thoroughly decentralized network it makes it really difficult for any party to have an inordinate amount of control whether that is a client like say um prism couldn't have control over the network if they had a minority client if we were all at 30 um whether it's geography like a nation-state actor couldn't take control over the network if if we're spread thinly around the world an operating system vulnerability if

someone was able to hack you know windows systems or linux systems if if they were sufficiently decentralized across all of these these factors then it really increases the robustness of the network and it's that's kind of my obsession yeah absolutely and uh and it applies to layer zero as well the people involved and so you know here we have um i presume academic researchers i don't i don't actually know um but they um it's an open protocol we are transparent it's published it's out there anyone can come and look at it and publish their work and we talk about it openly and um as all part of the permissionless decentralized transparent

um ethos that we have so um i uh i think it's an enormous strength of of ethereum as a as a system as an ecosystem and a as a community thank you yeah uh we did talk about call 81. yes um i missed so i saw you were on this call the merge community call my uh that's when my headphones went out and i i literally spent 57 minutes trying to get my stupid headphones working yeah i caught that like the last three minutes um but apparently there was a lot of really interesting not new but confirming information out of that yeah is that yeah is that what your sense

yeah i i think so i didn't see the first two calls i haven't read the notes um this one was uh there was good information but it was a bit ad hoc so you know um you you need to sort of look at it all uh i mean not everyone will be interested in everything right so there's stuff which was about for for stakers whereas i think the focus is more on um adapt developers and and application hosters for this call so um yeah worth worth a look or if there are notes uh that appear if you are running an application you're concerned about how the merge will affect that then um definitely um you know have a look at it at 2x speed

and see what see what pings you definitely plan to uh to give it a back listen also dank shorting i plan to you just finished that right yeah i'd wait for the explainers before you get to spend too much time uh i really those those are explainers that i can i can pick up and and these are are great because they're just above my level and so that i can translate it then to people who are struggling to get to that level denver looks really exciting yeah are you going dude uh it's complicated i um

i hope so i don't know i i um my family is still very coveted conscious covert aware uh and um you know our three-year-old hasn't been exposed and hasn't been vaccinated so uh we have an ongoing debate about like the risk and is it worth you know the expense you know the risk expense not the monetary expense but i i'm sort of hoping to drop in maybe for a few hours um but we'll see yeah that'd be cool i am not there but very much um planning to be at devconnect in amsterdam in april so i am booked and looking forward to it so i have already committed not to go to

amsterdam um but i know eastaker is doing a ton there and so easttaker is is just putting a ton of great stuff together so i hope everyone can make it to that um i i think i'll be attending virtually but uh we're just not ready for international travel yet um given the age of our son and given uh kovitz so maybe in a year or two yeah yeah totally get it yeah yeah i think we've pretty much done the important news uh yeah um so i i thought i found this amazing yeah like so and i i was gonna say yeah i was gonna save the e4 discover that i'm

sorry the language discussion for the end but uh it's so funny people people judge so much about this content based on that name and i'm like come on listen to it evaluate the content and you decide but yeah our documentation team cheeky so-and-so has asked me to review their their documentation change and approve it in uh on github um for renaming ether one to e2 throughout either one i need to do throughout yes indeed all right i'm gonna stop sharing if i figure out how to use all right um i i had fun i that's it i think this has potential to go on um i

i hope that everyone learns as much from this as i do um you just you're such a great wealth of information well thanks it's appreciate it it's uh fun to talk you through definitely different format for me than writing this thing which is his own special kind of torture but this is much more fun appreciate it yeah i'm gonna see if i can