What's new in Eth2 - 3 June 2022 with Ben Edgington and superphiz

Transcript

hey everyone i'm superfiz and today is what is today june 6th uh 2022 uh here in the us is it 2022 where you are as well then it's 2022. sometimes feels like 1950 over here but it's uh it is indeed 2022. uh so about every fortnight uh ben and i get together and uh we try to uh walk through finn's newsletter what's new in eth2 uh this is the the june 3rd edition uh i think ben and i both have a good time doing it i learn a lot and i don't know what he takes away but he tolerates me so he might even get a mention in this one at the at the bottom

i read it uh so let's get through it um hitchhiker's guide to ethereum wow this is monumental uh this is you know sometimes you see something and you're just kind of envious and you think you know i wish i'd written that and and that that is quality um john charbonneau i i don't know him um at delphi digital uh i don't know what prompted in writing the article but it's a it's a very solid very lengthy deep dive into the future of ethereum um and all the different um technologies that that we are we're heading towards uh and it's very thorough entertaining accurate um yeah

nothing bad to say about it uh if you have an hour or two to spare you it would be well spent uh doing doing a deep dive into that i did the uh the bankless live stream and i liked it enough that i added it again to watch later because um it it is it's consumable in the bankless format but it is and you know i i feel like i took so much away from it that i'm gonna need to simmer on that and then do it again in a week just to uh to re-enjoy it for me it is go ahead yeah absolutely i mean there's a lot of information there and it's just such a pleasure to read things that are accurate right and and comprehensive um there's so much misinformation and

you know we don't help ourselves because we're always changing the roadmap and uh this stuff is complicated but you know john has done a um a terrific job of really understanding and presenting uh very accurately uh all of the information do you ever have that experience when you're like wow this person really gets it and that's yeah that's the fun thing yeah so this this guy um just really gets it i i had the same uh in regards to bankless i had the same feeling with hal press um a couple weeks ago like wow he just really gets it he gets all the things that i'm thinking that i i couldn't articulate that well so that's a don't miss her the big reorg it seems like a long time

ago now it does yeah that was right at the start of the last uh kind of cycle of uh new and eth2 stuff so uh yeah and the bottom line is really you know nothing to see here i mean it's kind of kind of was an event at the time caused a stir uh have we have we designed a consensus mechanism that's fundamentally unstable are we facing you know people's transactions getting lost in real blocks after the merge and um martin koppelman of um of gnosis was a little bit shocked surprised by it but i think you know digging down into the root cause it was it was just a thing um we're it's is in the process of being fixed in fact

this this event this reorg where you know seven blocks got kicked out of the chain was actually an artifact of the way that we're fixing this thing um so unfortunate but uh yeah and some lessons learned about how we can we can coordinate better in future to avoid this i have to say that there are two in the last fortnight there are sort of two lessons learned from for the um core devs um both about being perhaps slightly naive about approaching the way we plan things and i think we need to think a little more adversarially i mean nothing bad happened but that's the ttt is that yeah so the robson ttd thing uh which we'll come to in a moment was one and this re-org thing i the the background is that we're rolling out

a fix to an issue with the fork choice that was discovered some months ago um but we we didn't plan to do a tightly coordinated fix client teams were allowed to um roll it out as and when people could upgrade their clients as and when and when we had half the network that was fixed and half that was not fixed then there was this um conflict um which uh which led to this this re-org uh lesson learned we we should have coordinated it and basically done coordinated releases and got everybody to upgrade like like when we do a hard fork um yeah i'd say this is kind of you know the the biggest issue that the beacon chain has experienced in you know since december 2020 which is

you know 16 months 18 months um and so you know 80 seconds of of missed blocks is not not not bad yeah taku took down 10 to validators for an hour i mean that was quite big i'll put my hand up for that but yeah i mean in the grand scheme of things those those two instances are trivial i mean there really has been uh almost nothing wrong with it for and it's been running for what year and a half um yeah it's it's true i love barnaby's right up it uh i yeah between the the writing that i could interpret and uh the wonderful graphics i was really impressed so thanks barnaby for

putting that i i loved the slow and sequential progression i was like oh wow i i get this that's pretty cool yeah yeah thanks by the way i'm thinking about fork choice stuff can can be tricky um but this this is great and uh he was able to access the data to make it accurate which was which is which is excellent uh let's talk about the deposit queue yeah and this is just a nicety uh that i observed i think you posted the link uh initially a physics on east take a reddit um and um so we have this limit so you can put down deposits into the deposit contract if you want to onboard as a validator

and you can do as many as you want at any time but they then get processed into the beacon chain and then they are activated and they're activated rather slowly so when the beacon train started it was four deposits per epoch which is 900 per day um and then when we reached 250 000 validators 260 000 validators then it ramped up to five per epoch so 1050 per day 1150 per day maybe that's pretty impressive and then uh when we um reached oh yeah sorry that was that's 327 000 yeah and then we we ramped up

again to six per epoch so it's like um uh whatever 900 times one and a half is 1 350 per day now so the rates so the background is so the rate at which we're able to onboard validators has increased and it increases stepwise um and what we want to avoid here is changing the validator set too rapidly if you've got a bigger validator set you can take on more so without changing it in percentage terms too too rapidly so we only want you know uh one percent or one or two percent of the validator set to change per day um just to keep things stable enough that the consensus mechanism is uh is stable so

that that's the reason for it unless the validator sets grows you can onboard at a higher rate um and the same when we get exits implemented uh well exits are implemented but when with you know you can't exit but nobody does because there's no withdrawals right but the same mechanism will apply to exits so they're rate limited at the same rate the home takeaway for me is uh is it's just an improved ux like more validators mean that in general we can onboard more quickly because you know as a home staker it's like let's get to the beacon chain let's look at the queue and how long do i have to wait so it's kind of neat to get a little get that a little more quickly yeah i uh i actually spun up a new validator um and it was my first experience of actually

being in the queue because i i got genesis validators but i thought okay pre-merge with us that's your right yeah when i say i i mean yes obviously thank you thank you for the reminder mr taxman um yeah so yes my wife spun up a new validator and it was uh interesting that uh experience being in the queue because you know you put your deposit down and it just vanishes right you basically get no confirmation and it took five or six days to come through and become active and uh even though you know in principle i know what i'm doing and i know how to find out the data is still quite anxious time yeah that is um one of the most common

you know people come to eat stick and they're like i deposited and it's gone i need to know something and i'm like hey i'm sorry it's it worked but you're not gonna know anything yeah and we're talking about post merge tightening that up because there's no um i mean the time in the deposit queue the time in the activation queue will still be a factor but the time between making a deposit and the chain seeing that deposit uh currently is about at 16 hours to a day but uh we can tighten that up to potentially even one slot i mean one block oh wow which so yeah we've been talking about that on some recent meetings um because we don't have this issue where we need some follow distance between uh and this voting mechanism that's

right um i guess djr released spec version 1.2.0 release candidate one snappily named it has got a proper name bobstone yes the roxton beacon chain release yeah i mean the idea is that we do these beacon chain merges with the candidate spec for the for the real thing itself so we are not expecting any substantive change between uh this release and the actual merge um spec release uh subject to finding

anything nasty you know if something nasty shows up we'll have to make some changes but this is a genuine uh release candidate it's funny i used to get so excited about the spec releases but now i'm kind of like i i know what's gonna happen like it's it's not gonna be anything uh too different uh let's see robson beacon chain yeah this was the other the kind of fun event i it was an exciting morning for me i it's just uh i was like are we gonna we can't go live on the call now because the beacon chain hasn't launched yet so it doesn't matter if we had ddd yeah yeah that was a bit of a planning fumble i th the graphs i've seen um recently of the hash rate on on roxton

yeah um i forget what it is that's it that one yes that that spike right so background for those who are catching up you know robson's got very low hash rate there's no reward for mining it so people couple of people do it out of the goodness of their hearts just to kind of keep it ticking over a mischievous maybe malicious miner came along and added a huge amount of hash rate to roxton um which accelerated the ttd so you can see the total difficulty on the top graph here and it just went went vertical so we had planned before this event we had advertised a number uh a terminal total difficulty which would do the merge event we just blasted through that due to this extra

hash rate um even before the beacon chain for oxton had been deployed so it was all a shambles but um we had to scramble to to fix that um would it be a problem for maynet no i mean nobody can increase the hash rate on mainnet by you know 10x 100x i mean 2x would be monumental that would somebody able to do that could do a 50 attack on um 51 attack on mainnet so you know it's they might somebody might be able to change your hash rate on mainnet by one or two percent but uh and even that would only usher the merge in more quickly it's not like they would uh cause any damage they would just get us there more quickly so yeah fractionally more quickly so yeah we we

had a scramble to sort of um just get everybody to override the ttd and and set it after we we then said it after the new beacon chain robston beacon chain had launched and all the various forks that happened and so on so that's been advertised it's out there now and uh we expect to hit it all being well on wednesday somebody's got to do a bit of mining i think the eef said they would um rent some hash power to uh um to usher it in to hit the eighth so look forward to that i actually uh put my one of my just one of my cards i don't have many cards now but on a nice hash so that i could be part of the rent cycle if i could have uh dm'd the the minor i

would have just asked them to wait until after the beacon chain launch so that we i know not that that would be good but it would certainly be more fun yeah yeah um interesting there's an exchange of messages in the in the block uh graffiti which was um uh which is interesting i i had forgotten about that i i think i saw that at a really weird hour and then forgot to to really process it yeah i think if you click on the mischievous up above then i think i linked to tim baker's tweet there um where he uh yeah this is this is from the miner oh that's great i had almost forgot that entirely thanks

for sharing uh all right uh so let's see robsten has undergone bellatrix fork now that i i wonder i know for el that's a non-event is that it's also a non-event for cl like there's nothing nothing observable changing it's just it's just like a more like a spec change uh yeah so there is some spec change um so we change some data structures to so that we can uh accommodate block headers from the eth1 side we can we can uh accommodate all that all of the extra information we need and it implements the new apis um and makes the engine api

active so that's uh what's going on on the consensus side um so it is it is a proper hard fork because we do change some internals and it seemed to go smoothly so that's uh that's good great everyone using roxton will need to update their ttd now i'm i'm actually using uh yorkdown uh thorsten's e2 docker and i think he had it pre-configured so either he had it pre-configured or i'm going to miss the merge it'll be funny either way let's see ah the official announcement of roxton

merge always nice to see yeah that's the original one is there an updated one uh yeah just with the updated ttd i think uh oh yes okay it was announced yeah uh yeah okay uh also remy and yorick uh showed how to set up a robson note on each docker i i used that one uh and uh remy's non-docker guide if you haven't checked it out good stuff yeah all good stuff deep docker is literally like five clicks the thing i always struggle with is is finding where he stores private keys because they're in inside uh the github repo in a hidden directory but uh once i once i remember where those are

everything is is pudding jell-o smooth i don't know i'm just saying random words main shadow fork six yeah and these things um you know were initially a very big deal we all got together in um amsterdam to do number three i think it was and we're all in the same room and it felt like a huge event uh this one passed by almost without comment i mean nobody bothered tweeting it and so on it happened last week um parry and marius are kind of running these things every uh every week um so just you know to remind uh those not familiar with maynet shadow forks basically

sync up a merge enabled eth1 mode on the normal ethereum network so it's got the whole ethereum state and then we pair that with um beacon clients that are merged ready and then at a given difficulty on the ethereum network um they do the merge so it's only you know a few nodes is a couple of dozen nodes on the ethereum network but they fork off the proof of work ethereum and join the proof of stake ethereum with the entire ethereum state so it's basically doing the merge itself is is effectively the same thing um to boot all of the transactions that are

on the ethereum mainnet are relayed across to this merged proof of stake network so it's a very realistic test and uh and we can see what problems arise uh and they're getting fewer and fewer you can track the kind of the way the charts go greener over the um over the weeks um yeah so various things tracking there um you can see how eragon has improved um is now getting there with some consensus clients what we're discovering in the latest ones is just a couple of incompatibilities um it seems that baisu and aragon are

having a little bit of trouble proposing non-empty blocks post-merge and uh eragon with lighthouse and teku seems to be some sort of incompatibility but otherwise basically everything is good i don't think these are serious bugs these are these are minor issues that need to be resolved yeah it's interesting um i know that there's a high desire uh i'm gonna pull this up to get um hold on one second to get uh aragon and baysu nethermind more of the market share and i'm eager to do that but then you know i look at this kind of data which is fine like i

know they're making progress but i'm i'm personally not not sold on you know full parody yet i kind of want oh that's kind of a pun i want to um i want to sort of let them harden the same way that guests are that's one of the reasons i haven't quite panicked about execution diversity yet yeah it's happening um so one of the things that uh is happening in the background is uh the ethereum foundation has this test framework called hive that you can plug different um ethereum clients into uh and test their compatibility uh and it and it's rules-based so you write you write a set of tests based on on rules and there's a sort of scripting

language that that it uses uh there are about 110 different tests in there at the moment um so uh geth i here on the grapevine is committed to passing all of the tests by the end of this week and the other clients are working on it as well so that's that will formally when they when they pass those tests and they formally meet the spec um some are still failing just and i think on minor things um and then and then the shadow falls should go through smoothly for all that's i i don't want people to think that i'm not interested in execution later diversity i'm i'm extremely committed to it and i i plan to make it a a hearty focus in the next few months

um but i i do think like we have a little bit of time we it's anyway just my thoughts kiln yeah uh this surprised me people have risen to the uh challenge um so parry um tweeted that kiln is processing more transactions than mainnet which is incredible because mainnet is basically full i mean so yeah uh astonishing that is yeah that's a great test people have really stepped up uh to testing the merge i can't wait for the nfts like that's those are you know the big drivers let's let's get some uh non-tangibles

um let's see prototype sharding i i was curious about some of this eip4844 so this this is about this is um a step towards putting data availability uh in ethereum making ethereum a data availability layer so this is all part of the roll-up centric road map big picture and you know what what we know now um that we perhaps didn't know five years ago is that execution is not the main thing that the base chain needs to do that's useful but the main thing is data availability to guarantee the ordering of data and that is available

to applications on the network for it for a period of time and that it can't be withheld by uh by malicious miners or or parties on the network so we want to move eventually to having massive data availability within uh the ethereum based slayer so that's full bank sharding with um proposer builder separation so you have specialist block builders building big blocks with a lot of data available and the point of having this data is that that's available very cheaply to to roll ups so roll-ups are where all the uh the action happens on the network and they use the base layer to to feed them the data that they they need

and that's all they need so yeah um proto-dank sharding is a is a step towards full uh dank sharding full data availability layer um it implements some of the cryptographic primitives and some of the data structures that we need but it doesn't do the proposer builder split thing and it doesn't do full data availability sampling the point of the data availability sampling is that you can reconstruct blocks if bits of them get withheld or are um lost somewhere on the network then they can be rebuilt from the bits of bits that remain and then there's there's this whole methodology around that which uh

john in the article that we looked at i was thinking which is terrific um so yeah so eip484 is a step towards that what's practical what's it mean for you and me it means cheaper roll ups it means um yeah is this is this the thing they were talking about this uh several months ago and and debated doing 4844 uh before the merge but they decided to delay until afterwards that is that the maybe maybe we've talked about so many things yeah i know i don't keep up either i i guess this isn't dependent on the merge uh it's it can be honest okay that's that's the only thing that i

wanted to make sure was right in my head yeah that's that's correct um it's i don't think there's anything there in their merch merch dependent okay staking so danny posts about the risks of lsd which yes yeah he's trolling right liquid staking derivatives indeed yeah what do you make of that i mean uh you're probably more familiar with that world than i am really yeah you know i i have i i took some very strong positions and uh while well i it's not that i would back down from

those positions i'm also less inclined to be i i'm more willing to step back and and look for other solutions um i i do believe that a majority lsd is dangerous for the network um but i also recognize the need for a lot of voices to come together and a lot of ideas and minds especially you know justin drake's research proposal is fantastic if nothing else i'm very happy to see lead researchers standing up for the health of the chain that that to me is the thing that you're like we're not going to let the chain sail down uh in for profit we're going

to uh to take care of the thing that we work so hard to build yeah and it's a really interesting interplay between sort of in protocol and outside protocol mechanisms i mean i i i get the feeling that you know um we post things like uh you know you've been through this right over the last uh uh months you post look guys we need to decentralize this and everything and and then you know the sort of hardcore bitcoin has start pointing the finger and saying you know haha it's ethereum centralized and yeah um will always be um and you've just proved it by begging everyone to decentralize but it's it's an interesting trade-off that what they don't get is that all of these

networks are fundamentally social consensus we try as best as we can to build in crypto economic uh mechanisms into the protocol which nudge people which you know encourage people to do the right thing but fundamentally it's about the people and people make decisions and if they value decentralization you know that they will tend towards being more decentralized and make those decisions so i think it is valuable very valuable to be having these conversations and pointing out the risks and uh um because ultimately the degree of centralization is designed decided by the people deploying validators um yeah

i i actually i had this kind of week-long thought epiphany that strength is acknowledging weakness um and when when you look at other chains who refuse to acknowledge weakness they just try to cover it up um they're not really strong at all the strength we have is saying hey we have a problem can we talk about this can we can we do something to harden this um that that really is what i believe the strength of the ethereum community is is that willingness to acknowledge weakness and if we start hiding that if we start denying it uh then i will you know i'll feel concerned more concerned about our health than

acknowledging that we we need to grow so yeah yeah i think and this this is one of the things that i'm slowly getting kind of over the years the the importance of the social layer the over and above things like on-chain governance and things like that which uh we we have never done as ethereum um and and for good reason i think uh it's it's really interesting it it took me a while to come to that because i you know i heard from metallic not be a fan of ancient governance at first and i was like what do you mean like it's it's the best form of governance and and slowly i've evolved uh maybe not slow like i've i guess i've always had an inkling of that but uh social layer governance well like the chain really is just a manifestation

of who we are so why we have to uh we have to continue to steward that so yeah yeah same yeah great explainers i'd love to see these pioneered by teku so adrian i i i love that he so he's obviously uh championing checkpoint safety at checkpoint sync lately um i i love to see people standing up for something they're passionate about yeah um and uh i love adrian's engagement with all this as well he just kind of writes his stuff he doesn't uh you know this is on his blog and it's his his thing it's not part of you know

his teku dev work so it's uh terrific um yeah and and you know to be fair the other clients have all implemented some form of this now as well so it's no longer only toku the question is um you want to sync your beacon node very quickly you can just get a state from somewhere and start from that state and then sync up you know a couple of dozen blocks to the the head of the chain so it's virtually instant uh under a minute to fully sink a beacon node but how can you do that safely you know um you can download your state from infuria but that involves trusting in fury that they're on the right chain or you can you know ask your friend for a state

well that involves trusting your friend's state um and adrian explores the kind of trade-offs in this little article uh here we we really need better mechanisms uh to cross-reference i mean ideally you just take the root of your state and you just can check that on beacon chain you can check it on etherscan you can check it on um the ef website you can you know check it on infuria uh just get a half a dozen and you know check it on twitter um just get a half a dozen cross references if they all agree then you're good to go he he really uh helped help me realize recognize the value of checkpoint sync over sinking from um from root i guess uh so i it and that

was a difficult concept because you but i i really appreciate uh sticking with it i improve work people are very very wedded to the sync from genesis um mindset which you know improve a work which is unambiguous about which chain you you follow um sinking from genesis is is the gold standard but in proof of stake it is uh dangerous um and we you know we need to change the mindset sinking from genesis is a bad thing to do is not somehow the best but difficult is actually bad do not sync from genesis unless you've got some sort of recent checkpoint to compare your sync with so and if you've got a recent checkpoint to compare your thing you might as well

just start from that checkpoint um it's it's a it's a really good insight but there is a big mindset shift needed in kind of uh you know um blockchain uh uh old guard um here to to come to this this is weak subjectivity for the technical name finding those things that are counter-intuitive and challenging those thoughts and correcting them that's kind of a fun experience yeah yeah yeah and you know again the the hardcore proof work people will we'll take a dig at this but it's uh uh yeah vitalik's got an old article on how i came to love weak subjectivity and it's it's good it's a good justification yeah i'm i'm kind of over um

over having concerns about what hardcore work people say anymore like i don't care like i i'm done like i've i've moved on you can you can keep whining all you want but we're moving on uh indeed indeed where do you want to head next uh yeah we can do uncommon cause merge podcasts do you listen to that um i did so yeah two parts uh podcast uh this is hazu the first part was danny and tim um and second part uh stephane um from flashbacks joined them and yeah they're both just just great conversations i really really good to listen to um

quality conversation between very knowledgeable people about really interesting topics well structured well led by um yeah i really recommend those two um uh podcasts yeah we're about to be in the car for an hour and a half so i have a really strong feeling that my my family will love to listen to this like don't you imagine they'll love it i'm sure i'm sure they will sleep soundly uh i i am very excited to check it out i just haven't made it a priority yet some of those uh i have to carve out mental space sometimes to consume something like this like you can't just do it on the fly you have to say okay i'm going to commit i can do this and then really engage so i look

forward to yeah i listen i listen to these things while i go out walking and i like the um the longer form stuff because you can deal with the nuance right a lot of this stuff is not simple especially so this one discusses mev uh and lido both of which are deep and and nuanced topics with multiple viewpoints to trade off against each other and there's no sort of yes and no answers um and and you need you need the long form really to get the uh the full understanding i'm committing to that today i'm excited now uh my family will will learn lots so bankless verge of the merge i didn't

hear that either did you get yeah this is this caused a bit of excitement because uh preston and oh yeah got it now i know what it is preston and justin drake both said yeah we're gonna we're gonna target the merge in august and uh it's caused a huge stir on on twitter um so they finally the video is up so you can watch it for yourself but yeah it's a great discussion um great yeah i love to see folks like a lot of times it is it's the same with this you can look at the stage and say aha this is going to be great i don't i don't care what they're talking about it's going to be interesting exactly validator anonymity i got curious about

that afterwards i was like because you know we does this so i will not be able to process that really quickly but um does it have any you go ahead uh yeah and so this is um one of the things that we've sort of known about for a long time and hasn't been a problem yet but you know is one of those potential issues on the network in that if you know the ip address of a particular validator or a particular an organization running you know a thousand validators on on one beacon node potentially you can selectively um mount dos attack denial of service on that ip address

and uh take those validators down which gives you possibilities for manipulating the network in different ways um you know censorship or uh yeah manipulating the randall or um gaining mev for yourself or or whatever so uh so in principle knowing people's knowing the ip address of people's validators is is bad um we've got you know people some people stake through vpn um it's it adds latency so it reduces your accuracy you can publish through multiple beacon nodes so that you've got multiple ip addresses which you know makes it much harder to discover uh you and harder to um mount denial a service attack against

three ips stuff like that but you know the gold standard is to do um uh i forget that forget the name of a single secret leader election whereby the network doesn't know who the block proposer is until the actual slot um and only the only the actual proposal knows that it's their turn so suddenly i find out oh it's my turn to to propose next slot uh and the rest of the network doesn't know that but afterwards i can i can prove that it was my turn and i acted correctly um that doesn't actually give anonymity though that's just a workaround right or a different solving the problem correct correct so that that sidesteps the issue of being able to selectively uh mount a dos

attack because the attacker doesn't know which of the you know 10 000 ip addresses on the network is the one the right one to hit yeah so um so we have this single secret leader election idea vitalik there are a lot of different approaches have been proposed to this involving snarks and other mechanisms but vitalik proposed a mechanism um sort of using some um the the swap or not shuffle that we already used in protocol which looked interesting uh somebody has done an analysis on it and found that it it in in the form that vitalik proposed it is not good enough um but is that yeah

does that anonymize ips or is that you said that's another in protocol solution so that's another single secret leader election kind of um approach yeah i i do get curious about and i and hiding ips and i'm not even trying to say they were anymore but does that uh does this address um obfuscating ips or is it on protocol changes this would be in protocol hang on i'm thinking of the wrong article here this is um yeah sorry this is not uh this is not a

vitalik thing this is also yeah yeah that's um that's so that's the third paragraph here is vitalik swapper around single secret leader election uh this one is about actual validator anonymity um using dandelion network which is out of protocol um yeah uh but again it looks kind of impractical so we've got two we've got two things here which are going on around the same issue neither of which currently look practical so yeah i'm like when we get to this stage sorry okay it's in the open research questions uh bucket and yeah if any great minds will

want to come along and give it a look uh you know some sort of onion rooting or whatever it would be really really interesting yeah that's uh onion routine is kind of what i was thinking and i i just you wonder if that's possible but what you end up with is is a set uh well naively because i haven't really thought through it but you end up with a set of validators who are all in an onion network with each other and so even the exit addresses are uh validator nodes anyway um i i really loved mario havels uh this is my favorite toy lately uh i'm staring at it we already pulled it up once but uh where's it at oh no it's it's the bordell tool right

but i really love i've been i've been checking it you know four or five times a day just to see where we're at toward merge this has actually changed since the last time we looked at it it does it did have that wow there's a lot of different data here anyway just very interesting i i'm telling people you know that the target is noon i'm sorry 12 p.m uh june 8th which my local time is around 7 a.m or 9 a.m well if the current hash rate is .36 giga hashes a second and we need 34 then uh you better overclock your bolts just yeah it just takes one cannon right well i mean one

well-sized cannon to uh point and fire so we'll see i was interested in this uh nethermine's interested in pushing it back for three or four months and others want to keep the pressure on uh what you said here really got me right so this is around when merge and i think the you know let's let's assume roxanne goes well if roster doesn't go well then then all bets are off right so um assume robsten goes well uh then i think on friday at the all-called des meeting will be the um serious conversation about uh when merge um and question one is do we push

back the difficulty bomb um there have been some initial skirmishes in this uh initially at the all-court devs uh a week and a half ago um and also at a um merged testing call uh this time last week yeah the nevermind team are arguing we we should push back the the bond by three to four months for various reasons um partly around user experience on on ethereum um but also because you know i think they they feel that they're not ready um for merge at this time others you know uh the death team are pretty keen to get on with it i'm ricky

to get on with it i know that uh danny ryan preston uh sorry i should say i don't know what danny's view is on this um uh i say justin drake and preston very keen to get on with it on the other side so my kind of observation of the community with this distributed network of independent teams and the tendency is to to move at the pace of the slowest right um which is frustrating uh to say the least yeah not in a bad way like it's healthier sometimes and and there's i'm not detecting much of a sense of urgency um in you know maybe maybe i'm i i'm incorrect but yeah certainly

nevermind's team when they were presenting their arguments for delaying three to four months were like you know what's a few more weeks on proof of work and so on so which which for me is just horribly complacent i i think you know proof of work proof of waste we need to turn it off um at the earliest possible opportunity you know of another few weeks another few months on proof of work for me is is a tragedy yeah so uh yeah go on i i say i i spend a lot of my time defending ethereum like i guess publicly locally to people i know who only know about the damage that uh cryptocurrency does to the environment and

you know i i spent a lot of my time saying we are better than that we are moving beyond that ethereum is not part of of that you know uh energy expensive um kind of thing that you know of as cryptocurrency and so i feel like every day that we stay on proof of work we are uh especially when we when we can or probably could move forward safely um it's just difficult that i love your word scandal like it is it is a scandal we we can move beyond this we're so close uh yeah yeah and you know we shouldn't go too early we shouldn't do it unsafely but um

and i don't want to leave anyone behind either so and the answer to this i think is to have a sense of urgency right i mean this is not a this is not a sort of second or third most important thing we're doing with our clients the this is it there's you know um there was talk on one of the meetings about getting all the clients uh passing the hive tests we discussed by mid july i mean why is it going to take six weeks why don't we do it this week okay just put your whole team on it let's just do it this week um and um and so the question and that's that's that's how things get done in this community in this distributed community it's when we all get together and and focused uh you know our laser eyes on on the problem this is

when there's this sense of urgency this guy comes to mind just do it yeah too right um so you know i'm wondering where this sense of urgency comes from and you know either it comes from this difficulty bomb which is maybe too urgent you know it's starting to become a problem but um i'm i'm for leaving it in place and just getting the merch done and just keeping that that pressure on but if we do decide to to delay the difficulty bomb i think we need to keep the the heat on the devs we need to kind of not let ourselves off the hook for two three months i also i feel like we've learned so many times the the merge the rob's emerge is not going to go flawlessly like let's not

delude ourselves nothing ever goes flawlessly if there are going like we should just acknowledge that like if it going perfectly is is the exception rather than the rule um so i guess we just have to take that in stride yeah go it goes well yeah perfect well unrealistic this may be too much hassle but uh goes well enough um if we survive with you know one implementation running maybe that's [Laughter] so yeah i mean my my suggestion and we'll we'll see if it's uh how it goes um maybe raise it on the awkward desk call it is um if we do delay the bomb uh

then let's set a date let's break with tradition and yeah and we saw that with the beacon chain launch um it was only when uh danny came out he you know he did a bit of due diligence with the dev teams and said can you be ready for such such a date then announced publicly we we do genesis first of december and at that point you know if we uh everyone pulled together to meet that date um that final consensus is difficult yeah yeah i'd like to see that again though i just sort of recognized this trend line of uh what i would describe as community willingness to accept a uh slow block times and i think right now we're no longer willing to accept slow

block times but it is what it is yeah i'm ready indeed i'm ready for 80 second blocks give them to me 80 second blocks yeah i mean the kind of flip side is there are two potentially negative consequences if block time goes too long that that becomes a hazard for the merge because we're in a regime that we oh that's right are not used to now roxton has been interesting for that it's already because robson's had very variable block time especially with this sort of huge um extra mining increased bl uh decreased block times rapidly and then when the miner switched their rig off it block times um increased massively this revealed a problem with voting on the eth1 chain and all of that onboarding

validator stuff and the other thing is that you know if let's say block times double then uh a response to that is to increase the block gas limit so that the gas per second is constant which is okay for a while um but at to having bigger blocks and we just again we haven't we're not testing in in that regime because we don't anticipate that so um it has another dimension of uncertainty it does end up uh getting us into risky areas and i think that we are too mature of a chain to get you know even a little risky we need to be safe i acknowledge that all right

[Music] just down here i saw steering i'm a huge fan of steerium yeah yeah tell me i sort of keep and i you know i see the twitter feed and so so on but i haven't really done a deep dive into this um yeah what's your take on it oh i love i love sterium i love the team um you know it's funny we spent so much time saying we need a single click installer staking whatever and esterium is essentially that um you set up a linux box and you connect to it remotely from a windows box and manage everything through the steering interface uh it's very well designed they even have they've implemented uh jim's uh

vouch and i don't know if dirk but at least vouch um so it's very well set up or very well it's just i mean you can look at the interface and see that it's just really nice it's a great way to and this is obviously the beta i've never used this interface myself but the the current version is uh very user friendly um i'd really recommend it for anyone who need who wants to administer a linux staking setup from a windows box so yeah i'm a fan yeah interesting yeah yeah i should take a look um all right anything else you want hey uh i saw my name there we go okay yeah we're just

shielding our thing yeah i do want to talk about the uh merge call series that uh eat stickers working on and you can find sources for that for in two places and our restake is gone type it incorrectly anvedica has done this great blog post about what we know about the merge staking series because so many of the details are still kind of nebulous we can really only see robston right now um and around june 8th uh we're going to do robston we have a a youtube placeholder so you can go

ahead and click remind uh i guess it's a remind me or schedule this or something and you'll be able to get a notification when this video goes live all right even if it's several hours of drift from the well yeah because it's not it's not going to happen when it's supposed to right it never does so you've assumed uh midday gmt um but it could be hours before or after yeah yeah i i don't think anyone i guess it all depends on when people turn their miners on um i i have seen we have some really great devops people i'm thinking of perry but i don't know who's running uh who's running this nice hash setup but uh whoever takes that charge is likely

uh take a lot of pride in hitting a target and so we'll see right yeah that'll be fun i'll clear out my my wednesday lunchtime in anticipation maybe wednesday all day ben i i always love walking through the newsletter for you with you for you i don't do it for you you do it for me uh it has been a great time and uh i look forward to seeing you again when i say fortnightly or no uh in a four nights yes in a fortnight in a fortnight i'm learning i've really uh tried my goal in life is to use that word is to learn shakespearean english

absolutely thanks everyone and uh we'll see you in a couple weeks