What's new in Eth2 - 17 June 2022 with Ben Edgington and superphiz

Transcript

hey everyone i'm super fizz i'm here with my good friend ben edgington how are you ben hey fizz i am very well thank you beautiful sunny day here in england sunny day here too what time is it for you it is 5 p.m in the afternoon okay and it's uh 11 p.m for me here it's june 20th and we're about to talk about the newsletter that you put out uh june 17th um how are you how are you feeling about the newsletter is it is it i know a couple weeks ago you're like i don't know if i how long yeah i continually wrestle with you know what does the future look like do i see myself doing this in a year's time two years time five years time one month's

time uh and and all of that all i can commit to for now is you know carrying on for a few more few more issues and we'll see how it goes um it still does the job which was its original intention which was to keep me focused on ethereum core dev stuff so that i you know for my personal benefit and my team's benefit i know basically everything that i i feel is important like you know forces me to interact with it uh listen to the podcast read the documents and and you know find something sensible to write about so in that sense it is doing its job um people seem to continue to find value in it and uh appreciate it which

which is great lovely by-product as far as i'm concerned uh i really appreciate that but yeah we'll see yeah i i've noticed this kind of thing um when it comes to public service especially like unrewarded public service like the the personal benefit the personal reward only goes so far and then at some point you're like why am i still doing this but i i it is very valuable appreciate it hey before we get going fizz i need to introduce you to my new friend this is uh my my teku beer glass i saw that i didn't understand so apparently the finest beer glass in the world

is it i guess i'm curious when i looked at it it was it had the ethereum logo in the name i know it was crazy isn't it i i don't know which came first that's what i'm trying to understand yeah it's embossed on the base this little it says tecu and it's got like the little e sign just for three bars for the for the e uh yeah it's insane um yeah this glass has been around like 20 years or something and uh it's um i'm not drinking beer wow that's the finest uh beer glass in the world take a three three point zero that is i because i i looked at that i was like what which came first i it just i really i was racking my brain about it like trying to make sense of it of course like i could have probably gone to the

internet to find out but there we go the internet it knows congratulations i don't believe everything you read thank you uh okay so what's new in eth2 17 june 2022 great glacier uh delaying the difficulty bomb um it's going to hit june 29th i know you have thoughts about that uh yeah i mean it's happening right now i've written a little bit about it uh further down um so update your eth1 nodes anyone running an ethereum uh well there's still eth1 nodes until the merge at which point they become execution nodes um anyone running an ethereum client you need to

uh upgrade because otherwise you'll be forked off of the main chain so it's only a tiny tiny update it just changes kind of one number in the code base and just pushes out the difficulty bomb for uh i think 100 days or so which takes us so we'll be in the same position at the end of september as we are today with block times being about 14 and a half seconds so normally they're 13 seconds on on mainnet they average out to that um so uh they're 10 longer now than than they would normally be that's expect if we don't defuse a bomb that will grow rapidly and we'd be looking by the end of september at something like 30 second block dimes or something

so this just delays that process by three and a half months and buys us a little more space before the um before the network becomes too congested and thankfully uh geth is really easy to update in general uh most people will have installed it from the ethereum ppa which is simply a pseudo-app disk upgrade gas and it pulls it down which led me to like realizing i i have this epiphany every like six months why aren't the consensus layer clients doing that um but that's that's another discussion yeah it's a good question uh we can look into it um yeah uh it's just another thing right but yeah if you if

you want to raise an issue on the teku repo and the other clients as well i guess um make it so yeah i think i will do that today thanks for the encouragement so rhapsody has merged all right yeah that was uh success you were there the the longest ever wait the longest game of the song today um that was going on forever that cool so the the issue was that um the idea was that because roxton hash rate is very unpredictable normally it's very very low right and but you never know when somebody might step in and and really ramp it up so we saw that uh earlier on in it and it rather disrupted things so the the idea was we put rather

high terminal total difficulty on and the goal was that the ethereum foundation guys would load up with hash rate in the hours before um the target time and just kind of bring it in just by loading up the hash rate accelerating this thing and they from what i understand they put it on a too smaller box and it just kept you know falling over in different ways and so this hash rate was really intermittent and and just wasn't doing the job so yeah we expected it around at i think 2 p.m my time and then it then it was looking like 3 p.m then 4 and then it just kept going out um so yeah but uh you filled the time and uh it was um it was fun to uh spend spend the afternoon with you guys

yeah i had a lot of fun it i it is obviously a little stressful um you know trying to keep the energy up but yeah i'm so appreciative of jonathan mann um and just everyone for kind of hanging out and enjoying the process because that that is the fun of it that it's the fun of doing it together um so yeah i had i had a blast yeah yeah and we've got a few more to go of these uh these get together so looking forward to it can't wait uh so perry always has great summaries um i was i don't see danny's linked here but i was a big fan of danny's um big success tweet uh that's the one yeah just below so uh we can look at that in

a moment that's the absolutely [Laughter] so yeah so perry tells it how it is um there there were some bumps uh on the way um we didn't you know it wasn't auto perfect smooth sailing for every body um process getting from proof of work to proof of stake in particular uh nimbus team just had a typo on their command line and um you know uh nine percent of the network was being run by them and when the merge came that they they dropped off so very easily rectified you know quick restart of their nodes uh and they were back i love this

the memes are good and i love nimbus as ascii art um i haven't i haven't seen it in production i guess it went through on the test net uh but i will look forward to seeing that i think there there needs to be some kind of ascii art competition now yeah yeah we've got a few contenders for techa we need to take a look and uh and choose one so yeah i always appreciate uh perry's really thorough um reflections on this and that's it's hard to know i say this during the call but during the call you know what we saw on the call was essentially a delay and that's you know that could be frustrating and people could interpret that as a problem um it really takes several hours afterward to know

um how successful or unsuccessful the event was yeah and this is one of the things i've tweeted about this in the past yeah what you want is a moment where you can just say yeah it's done and you know you can all start applauding and break out the champagne and all of that and it's never like that you know we've simulated that moment for some you know um publicity purposes you know like when we did the amphora thing there was a kind of point where somebody said yeah sure we'll clap it's like yeah i think it's good let's do it and you know i got the photo it just it feels like a little bit of a letdown during you know the the hour or so afterwards because you're you're always just saying has it worked has it not worked yeah what are the problems

why are we not seeing this or why are we seeing that and yeah i think we have to acknowledge our emotional side and people need that moment to say yay like we did it because if you if you don't have you know the ascii art screen saying that merge has happened you don't really your body never really appreciates the transition you end up you know a month later did that really happen or not i don't remember yeah and you know the banners will appear at the at the point when proof of work gets switched off we don't really know that proof of stake is fully functioning for you know quite some time afterwards i mean at a minimum like 13 minutes when we get finalize that that stuff celebrates

yeah yeah come on all right and this this is a great tweet from danny uh calling it a major success yeah so i i've seen all sorts of takes on this across twitter and i wanted to kind of call it out and say look it didn't go perfectly but it went well enough and um tim on his bankless episode which i've linked elsewhere yeah makes a point very well that the ethereum network is designed to be resilient right you know people are out there saying oh it has to go perfectly we've got to test this thing to death it's got to be until we've executed absolutely perfectly end times we can't do the merge but i actually i that's unrealistic that's never ever going to happen we've got you know thousands if not tens of thousands

of amateurs running this network right and they're gonna they're gonna mess things up right and you know um there are gonna be nodes that don't make it that aren't upgraded configured wrongly that all over all sorts of things are gonna go wrong but will we make it yeah of course we will it's designed to be resilient um the the ethereum network will survive 10 20 30 of the nodes offline uh with a you know maybe some small degradation you know we might miss some blocks or or whatever but it will keep on chugging and i want to be kind of out of this mindset where it has to be absolutely perfect now that's kind of the protocol has to be perfect the design has to be correct we don't want any edge cases that can be

attacked or or hacked or put us into some weird loop that's irrecoverable but when it comes to you know do people configure their nose correctly yeah we we're robust the theatrical production doesn't have to be perfect that's really what we're looking at because that is you know that is what the call is that is the event is the theatrical production of the merge because we still have plenty of time afterwards to uh you know reboot nodes or to change parameters um and get everything going perfectly yeah i i think that you're kind of right people are are hung on this idea of the theatrical production being perfect when

it's really a side show it's something that that you know we want to celebrate together but in the big scheme of ethereum it is it's it's not really that important it's just a chance to be together yeah and uh yeah as long as the chain keeps on chugging even if it's yeah mildly degraded for a couple of hours this this is fine it's you know nothing's going to break if it stops you know the worst case is it it stops so we we have some sort of fork and you know no blocks are produced or something like that um that that is a bad situation and you know d5 stops that's that's disaster we're pretty confident that this by this stage the number of tests we've done and the number of dress rehearsals um that's

that that's not gonna happen so uh yeah it was great so um the bomb has been pushed back uh yeah so i mean the big question is what does this mean about when merge um you know i speculated and i'm in conversations with with various uh parties around the place you know we're always talking about this um one way or another best view on this is that uh we we set this thing back to really start becoming a problem in mid to late september we don't want to do it again that would be a really bad signal

that would be troublesome you know everyone having to update their notes twice to to push back the bomb and that was agreed by all teams the 700 000 blocks was agreed by all teams as we can realistically do the merge you know subject to the caveats you know no major disaster blah blah but yeah on on our current understanding it's it's realistic to do the merge by uh mid-september uh there's been a commitment from all teams uh execution and consensus teams to to that delay so you know it's it's not a schedule for the merge nobody has said we will merge mid-september but the timing of the bomb delay is very suggestive uh and agree by all teams and

there's buy-in and so on so i i'm bullish september uh at the latest sure yeah let's you know we i i always uh celebrate this multi-implementation architecture um but it does end up becoming uh hurting cats it is i don't know i guess you have to accept those uh inherent slowdowns or the the difficulty in working together when you have when you have decentralization so i guess maybe that's a great thing maybe it's a good thing that that we're seeing uh people

pushing and pulling in different directions because that that is that's actual evidence of decentralization if everyone wanted the same thing all the time then we probably wouldn't really be decentralized yeah i think so um and yeah having so many people involved is um a lot of eyes on this and a lot of different approaches it really does test things well there's a danger that we just move at the speed of the slowest and at some point we need to say okay enough is enough you know the slowest you know you catch up later uh yeah we're doing this without you and that's what slightly worries me about not setting your time is we haven't set ourselves a criterion by which we say

okay this far and no more you know you're in and you're out and you know i had a sort of twitter um discussion with tim baker on this and i'm i'm more in favor of i think we should set a date we should say by x date we will have done the merge barring you know really unforeseen circumstances you know every every program i've ever managed uh in my in in the old world uh we we did a timetable it's just normal project management you do it to the timetable um but there we go um it's happening so much of the uh the consternation i've heard is at least you know on the us end user side is um you know we still don't

have execution client diversity and why don't we have that yet and from from my perspective for better or worse i'm not quite ready to promote um all four of the execution clients and i i also don't think that it's wise to to wait until they're all ready because that could be that could be a year that could be many years it may never happen uh and so i i do agree with you that the best thing we can do now is is set a target and move toward it uh carefully but yeah yeah and uh you know this time i i've been hardly encouraged by the growth and diversity of consensus clients so uh yeah

where there's a will there's a way it'll come absolutely yeah i agree with that sepolia let's move forward yeah yeah yeah so sepolia is one of the existing ethereum test nets it's a proof work test net um i don't know much about it it's not hugely used but we're taking the existing ethereum test nets and emerging them as dress rehearsals for the the main event um so in order in order to do this we need to set up a beacon chain to merge it with so that beacon chain went live a couple of hours ago and seems to be running fine you know we know how to do this now it's straightforward uh and the goal is i think to do the merge on the 29th for sepolia

target it rough roughly for that which is uh what next wednesday so uh yeah are we doing another um get together for that yeah we are actually um so we were kind of stalling i i hadn't heard the date i did miss the last implementers call um so i hadn't heard 29th of june yet until i read your newsletter but uh anthony cesano from the daily gui is going to be the primary host um and east aker's going to support him and obviously we're going to invite all of the same people um and i i think there's a fair chance he'll get lucky and not have to stall for five hours but uh it should be a ton of fun yeah excellent i'll join if i can uh yeah that would be great can't wait

uh so let's see i was thinking you know i really there's a part of me that wanted to ask to be a separate validator then i'm like do i really need more responsibility yeah maintenance shadow fork 7 ttd on wednesday the 22nd yeah these these are happening so this is where we take the current stator mainnet we put some ethereum one nodes which are merge ready sync them up with mainnet and we set a terminal total difficulty and they they just fork off um onto a specially built beacon chain uh and become proof of stake so we've got the whole state of mainnet you know absolutely everything

that's out there and we relay the transactions across as well so it is it is doing the mo it is actually doing the merge um we could just say you know 20 minutes later okay that's now the merge maintenance it could be it could be like that but uh really well not really because obviously yeah well i mean yes i know all the values on the all the values on the beacon chain of the staking so we'd have to do some kind of some kind of funkiness to to move it over but yeah it is it is absolutely the full mechanic of the merge with the full state and and everything so and number seven is is coming numbers

four five six went pretty well um we're getting that great very excited to watch it so i've been a bit big fan of this the merge readiness checklist yeah apparently i was thinking of actually uh the i was thinking of the merge right this is it uh the last time i looked at this it was a github uh or yeah that's right it's being published okay it's been published so this is live on the ethereum launch pad and is um and it's good yep um it's just the main things you need to check so what have we got um runner runner both clients so anyone running in

ethereum know today whether you're a a staker you need to run your own execution mode you know you might be using infuria as i am so i need to spin up my own local eth1 node or if you conversely if you're running you know for whatever reason for your dap or for your you know just your own um reasons uh and eat one node you will also need a beacon node in order to be able to sync the network not with validators but just run a beacon mode so um everybody will need to run both of these because they're kind of glued together uh post merge now i'm just gonna have a little rant here so this is often a digression that i really deplore the this appearance of the word layer everywhere this has become kind of the vernacular

now we talk about execution layer client no it's an execution client consensus layer client no it's a consensus client can we just all right yeah i get that skip the lay of it it's interesting yeah uh i you know it's funny i had never really thought about it and now that you say it i'm like oh that's that makes obvious sense once you process it for a moment yeah that's just one one of those little bees him up on it i don't know if i'm just inclined to agree with things you say but i'm like that makes really good sense say something i don't agree with just as a test and i'll be like ben you're full of crap we'll get to that the engine api uh yeah uh authen the authenticator is the is the point the main point here um so you've got your eth1 node you've got

your eth2 node or your execution gun your consensus client uh and they you point them at each other uh well basically you point the consensus client at the execution client but you also need to secure that connection so that this is a new process with in the old style you point your e you know consensus node at your eth1 client you don't that connection is unsecured you're just kind of reading state from the uh execution client so it doesn't need any security in this one you're actually sending things to the execution clients you have control over it so it needs to be an authenticated connection and and we're using this jwt mechanism so you need to generate a

random token and feed it to both clients so that they can talk to each other so a little little tricky um i bet a lot of people are going to get this wrong and this may be one of the major causes for drop-off when the um merge happens however that they should the clients should as soon as bellatrix is done it should start which is a fork before the merge on the consensus side the two clients should try talking to each other and if they're misconfigured then then you should get some loud warnings in the logs so it should tell you if it's well configured or misconfigured that's great i ought to confess that i have not i haven't played with this yet um i've been aware for a long time that it's

coming but i haven't uh i haven't played with it in my server yeah on the on you so you can check it out on the test now it's on mainnet for now uh beacon chain maintenance it's not useful but uh yeah on the test notes it's um you can you can try it out that's great uh so it won't it won't really be active on um may not until at until after bellatrix is that right correct yeah yeah yeah that's when the engine api kind of goes live which is the way that the consensus and execution clients speak to each other and then yeah this this one's really really important set fee recipient now the network's not at risk if you don't set your fee

recipients so everything will keep going only um by default any transaction fees you earn which is the minor tips will just get sent to a burn address unless you set this um in which case you you will so post merge you've got two sources of income or maybe three we'll come to that you've got your block rewards as currently um for producing blocks on the uh which go into your validator balance on the beacon chain and you have in addition um minor tips so transaction fees from ethereum transact transactions which go to a normal ethereum address and are immediately available to you so um you need to tell your validator client and

your beacon node what your uh fee recipient uh address is where you want those tips to go so it's like the um coinbase in in the blog if you're a miner you said a coin base which is where your block reward goes to it's basically the same thing and i i definitely took a few hours to set up my fee recipient even if i missed jwt i i wouldn't miss the recipient yeah yeah this is definitely an important one so there you go um when merge cool that's actually um i'm i'm glad we shared that and went over it because um this is something i'll definitely want to promote quite a bit it's probably one of the most valuable things for home stakers to be aware of right now yeah that's good stuff

mika labs report pretty impressive but i i saw the exact same thing that you saw the the focus on um sync times and definitely a worthwhile read uh i i definitely uh got a lot out of it but uh yeah go ahead what do you think yeah i so they've done good work i mean i know these guys uh barcelona supercomputer center leo has been around forever in the ethereum community and um you know they they've spent a lot of time and been very careful in their measuring and analysis and the situations they've set up and you know before the report they were in constant dialogue you know making sure that they were testing um correctly and and so on

they've taken a very responsible approach there's so the the the data is good um and you know the conclusion is that all the clients have their strengths and weaknesses they're all basically good um and you know you you can't really go wrong there's some useful data some feedback uh for client teams uh coming out of it in the written report they focus almost exclusively on sync times so that's when you start a node with a blank database it goes back to genesis it starts loading blocks and recreating the state and syncing up and this on mainnet this takes beacon chain main that this takes a few days um currently um and and they measure cpu usage memory usage bandwidth usage and so on in that

situation and my problem with that is clients basically spend nearly all of their lives not sinking right they're not measuring the right thing my my own node has spent maybe i know five minutes sinking in the last year and a half just when i upgraded and restarted you know it's just got to sync up a few blocks um it's spent you know a year and a half minus five minutes not doing what they've tested you know there are ways to optimize sync which we don't do in tekku some clients do you know you can but we we see it as pointless you know to spend time doing that we've spent the time rather eliminating the sync completely start from a checkpoint state

and you know you're up and running in two minutes without uh without having to do the the whole thing process um to be fair they they have measured some other stuff they haven't really included in the report the the plots the graphs that they uh produce under e32 client plots there um contain a lot more broad broader data about some tests that they've done so if you're interested in this stuff there's some good good things to explore there yeah it makes me uh makes me think of jim mcdonald's work like i love how or well i guess i i should refer to them as a testament um both attestant and amiga labs do these really deep dives into chain metrics and i i just love the

academic nature and like the the quality of their research is just it's fun to read yeah that's great um yeah and what i i was going to mention the what a testament have done is kind of looked at the problem from a different angle which is uh the reward earning capability of the clients and so they've done a lot of comparative studies between who builds the most valuable blocks who has the most um reliable attestations uh and so on uh and jim's talk from dev connect um at the staker event was was super and again whose conclusion is all clients have their strengths and weaknesses you you can't really miss out especially if you're on vouch which is a client sure kind of you know we should take a moment while

we're there uh jim mcdonald works with a testant uh and they develop a product called vouch that essentially allows you to run all of the beacon chain clients on one server assuming or well multiple i guess if you have that hardware configuration and it it reads the data from each of those to choose the best submission for the network and so using vouch um while it does have a higher hardware requirement it will always give you the the best network input that you could provide is that fair uh yeah um absolutely i i don't know if people will run it as a kind of uh in a homestaking

setup i mean if you want to throw hardware at it you can um you're basically running multiple beacon nodes yeah so more demanding but uh but if you want that extra you know 0.1 or 0.01 increase in in reward then then you know uh might be worth doing i i like to think of it and i know i'm an oddball for me the the reward is is um being highly engaged with the chain it's not it's not necessarily like i'm getting you know a millionth higher of an ether reward it is uh just being that much more intimate with chain operations like i'm i'm doing better you are a good citizen

i'm a strange dude congrats to wagyu yeah tell me about this i'm sure you're more familiar than i am on this one so they have been working on wagyu primarily from uh from my view is is remy and buddha um and i'm sure there are others involved but steakhouse is a an initiative that's that was born out of staker um and they develop um so they're essentially the the developer arm of e-staker they create tools to make staking more accessible whereas e-staker uh provides education to make staking more accessible

and so the wagyu installer is is a native gui installer uh manager for for staking and i they've been extremely careful and cautious about you know incremental development not over promising and i i'm just i'm really proud that they are that they're getting it delivered yeah i've seen it sort of uh being developed over the last year or so and uh it's terrific yeah yeah i i really ought to uh i really ought to test it more um and i know you know we'll get some more videos

about it but i'm really proud that east acres is primarily education focused but to really produce tools is is really you know it's a great way to give back to the community so i'm really happy about that mev boost i didn't really highlight anything here because you know i just i have not been that engaged with me it's coming dude um yeah if there's one thing to read or to watch in that section i would pick up the video preview of uh mev and mevoose for validators from remy eureka and uh led slash uh that is excellent um they it's it's kind of informal and they just talk

through the whole picture and then do a demo of what what it looks like to run a node with the the medfoot sidecar and yeah that's just quality um gives a really nice intro and doesn't gloss over the difficult stuff go it goes into it but gently uh and uh yeah liked it a lot that's great and i'll tell you um from eathstaker uh both remy and and york are our uh well and buddha also are our kind of mev experts and i have sort of just i've turned a blind eye to it i guess like i haven't really you're in denial you're in i am i don't right or wrong uh all right um

i think you know this guy somehow adrian sutton heard a few yeah he's a good dude um so yeah uh just going back to what we talked about earlier about not spending time on sinking from genesis um so adrian gave so ethereum engineering groups run by um um peter robinson who's one of my colleagues in consensus protocols has been for years he's in brisbane australia so um he runs he runs a group um and has technical talks on things there's adrian top left there peter's top right um and some other familiar faces uh so we put into tekku this checkpoint sink i

know a year year and a half ago we sort of pioneered it others have picked it up the the point is to sinks in proof of stake to sink safely from genesis you need to know a checkpoint on the right chain to make sure that when you've synced up you've ended up on the right chain and not been diverted onto a fake chain that somebody has done a time bandit attack on i forget what it's called long range attack that's the official name gone back in time and built an alternative chain based on keys that they've stolen or bought from somebody so if you need to know so you know one issue is you can sync up and find oops i'm on the wrong chain and then what do

you do you have to go back to the start and kind of hope that you get a better set of peers and end up on the right chain you know that's a problem uh and the other issue is yeah if you got this checkpoint of of the which is the state hash the state root hash well you might might as well just use a whole state and just think from that point onwards why why even bother going back to the beginning you're trusting this digesters they just use the whole state start from there and and forget about the the sink um and so that that's the premise of it so the trust model is is unchanged by checkpoint sync compared to a normal proof-of-stake trust model uh so uh adrian talks about that more more authoritatively than i do so i i just i love seeing somebody that has

a passion for something and i see that he's passionate about this and it's you know you you want to see someone who has this passion like keep working toward it and then slowly turn the tide until everyone gets it yeah yeah exactly and i think his passion is kind of a a reaction against yeah under proof of work the only safe thing to do is to is to sink from genesis right that's the gold standard uh and find the the heaviest chain but that doesn't work under proof of stake and so people keep throwing at us this uh oh you're not thinking for genesis thing um but you know that that's an error there's a category error so people yeah i'll be honest i i didn't really uh

fully accept the connection until his um devconnect call was it different yeah uh i'm sorry it's devconnect talk and then i was like oh well now that that really makes great sense uh so yeah i appreciate his work uh this is a reprise of the devconnect talk so if you've seen that you don't need to do this tim on bankless yeah a 100 100 tim uh very tin uh just great i love it i miss working every day with tim um he he's uh uh he's fantastic yeah i really like the guy um let's see tell me about your bankless experience my terrible terrible confession i

haven't heard it yet i prioritized him over you forgive me i'll get to it in the next day or two uh no i don't like it is really just you know i some of my my friends are like you know how did you feel about your talk um and you know i listen to other speakers and they use uh you know complex ideas and rich vocabulary uh and my talk is essentially hold on the big guy's coming we can play in the bonus room my talk is essentially like uh why decentralization is important why i fight for decentralization and

you know i i i even felt like i repeated myself um several times and you know uh it was good it was a good opportunity to get those thoughts out there and i really hope they're beneficial to people i really want people to value decentralization um but i also know you know a lot of folks already get it and that's that's okay with me like you get it you're not you're not going to have any epiphany listening to me talk about sexualization well bangla reaches a wide audience right and um uh yeah i've really come to appreciate them over the last couple of years when i've been listening to podcasts more um great look forward to it neat saying yep need saying and repeating uh yeah

i i love to talk about those concepts i i just i know that it's it's not it's not new for most of us i love ech uh their peep and eep series do you want to uh uh not much to say on these i mean the people need sessions are a gold mine and they they just come up frequently um they're doing uh pooja is doing like one or two weeks sometimes um and they're not sort of broadly advertised and so i you know i look at some of them and they've got like 50 or 100 views and i think this is tragic i mean they are so i always try to include them because they cover a huge range of what's happening in in ethereum

generally around eips and stuff that's coming down the pipe um and have you know quality speakers on them that you know the people actually doing the work um unlike me and it's terrific so uh really you know if you want to dig in to really go one level deeper into ethereum stuff then engage with these peep and eat um seriously the cat headers channels are linked to the playlist it is uh it is the opposite of fluff it is like uh pure mainline information so it's it's pretty cool yep yeah totally um yeah then eat prague prague looked like a lot of fun

i didn't make it there i haven't watched these but uh yeah um worth a look there good 20 or 30 minute talks um peter's one at the top is not you know particularly about um proof of stake or future of um big ideas about the future of ethereum but he does explain and coming back to your decentralization point we can crank up the transactions per second in ethereum you know we can quote scale if we want but um there are trade-offs to be made and he explores those trade-offs and what they are um and that's you know we we cap the tps to really there's that other

the other layer one chain uh that like takes us to the other extreme uh you know totally has cranked up transactions at at the expense of extremely heavy hardware um and you know there's evidence they have trouble even keeping up time so um uh yeah well i can think of more more than one that fits that category yeah that's why we don't say any names exactly and you know where do we draw the line well we we it is decentralization is the is yeah making it well and you know you kind of have to define these terms and you know all this right but it's uh it's

making it easy for anybody to spin up a note or as easy as possible the broadest possible participation in the network yeah i think that's where we'll win if by just increasing access um i haven't seen this twitter never i don't even recognize it no um yeah i just came up on my radar and seemed to be saying sensible things about um how uh rewards work post merge and um income and the different income streams that validators will have seem to be information rich so put it uh put it in there it's funny uh and i'm not even if i should bring this up but it's like a sexualized pfp and then like really deep

uh thoughtful information it just it's kind of a contrast cognitive dissonance let's see awkward devs so yeah we just sort of talked about that yeah very eventful two things yeah yeah tim described it as very eventful i mean you know i mean i can think like uh eip 999 when with the parity funds recovery you know that that was a bit more eventful i mean you know there's been plenty of drama over the years and i don't think this was particularly dramatic so i'm not quite sure why he felt that i'll tell you i i blocked out i was like um you know honey i'm gonna be you know locked in my room you know five

hours i can't do anything else i really expected to walk out of that call with a uh emerge date but that was obviously characteristically overly optimistic a phase so yeah i i have an issue on the all called devs calls that i get i don't speak very often over the years over the last six years i've maybe spoken four times at all called their schools and every time i've got hopelessly tongue-tied which is not like me at all i mean i'm usually reasonably fluent and i just say the most stupid things and it just all comes out wrong and they'll just feel like total idiot on these things and you know it's not all these are people i know and the people i speak to every week um you know i'm not intimidated particularly by

them or anything there's something about sense of occasion of being speaking on an awkward death school that just throws me for a loop and i sort of leather like an idiot very interesting i i yeah i i guess um i've heard you speak more on the implementer's call so i don't it's hard in my mind to shift the uh i guess i would say they all blend together in my mind so dank sharding you have your own section for dance shorting now well we had a breakout room meeting so this is in the this is in the uh regular calls thing so this will be a regular thing maybe once a month um

and it was flagged up at the um consensus uh local so 484 is the protodank charting so it's not the full thing it's what we might put into the fork immediately after the merge or upgrade i should say which basically increases block sizes but makes a big portion of the block space data which is not available to the evm so um there's no guarantee about storing that data forever we might store it for a month or a year or something but is bounded and you can use that data for roll-ups um and uh you can prove things about it to the evm using kcg commitments and so on so

all a bit technical but we if we want to get this into the upgrade after the merge and not delay withdrawals so currently the upgrade post mode upgrade is you know some tidying up withdrawals and proto dank charting eip484 are the main candidates for inclusion and yeah clearly withdrawals are sort of systemically important at this stage um if we don't want to delay withdrawals we need to get moving on 484 sap because it is complex and it's a big kind of quite a big re-engineering of stuff in um in protocol um i'm i'm tempted to sort of pitch for yeah do split it into do a quick

withdrawals upgrade maybe four months post merge and then maybe eight months afterwards to do this but we we're not talking about that now that that's that's me that's not you know cool so let's let's let's view that in the scope of ethereum time it's practically instant isn't it i have to acknowledge uh a lot of this is still over my head but what what i do think is a great resource for most people is going to be that john from delphi digital talk uh on bank list because he does uh he gives he gives an overview of this that that was far more consumable for me yeah yeah totally so i think i linked to that in last issue or the one before that so if you go back a couple of weeks

and what's your need to do you can you can find the links to that yeah yeah uh lido's digesteriums ninth episode uh and knows this beacon chain is really interesting so um they um used to be called x-die all my perhaps or most of my co-apps are on x die so i hope they don't mess us up um they um are planning to do the merge so there's a beacon chain and there's the xbox with the gnosis network i forget what they call it but basically the formerly x die and they're planning to do the merge on those two ahead of ethereum mainnet's merge

um slightly more challenging they have some oddities like a five second block time uh on the beacon chain which is um uh a bit challenging and uh some other bits and pieces but uh yeah teku acts as a validating note for that um is it techo in lighthouse is that are they uh yeah i think that's right um prism is work in progress prism got a grant from lydo to to do that um so yeah i have some interest not not technical interest but just fonzie interest in uh running one of those validators it looks like it's uh requires i believe it requires 32 gnosis coins and that's

about 3 700 right now just something fun to do maybe there's a yeah they're using m which is i believe that i believe em no is wrapped gnosis which is essential would be the same price as yeah i believe that makes sense i'm not i'm not too deep into it just yeah you can earn rewards well there we go it's a wrap that's fun good talking to you i'm going to shut off the recording and uh thanks for everybody for joining us we uh we'll see you again in a couple weeks yes sir look forward to it everybody