What's new in Eth2 - 29 July 2022 with Ben Edgington and superphiz

Transcript

foreign okay so Ben and I were in a conversation but I'm gonna do the intro and then we'll do the conversation just because it's fun that way uh so hey I'm super Fizz um I'm uh organizer with the e-staker community I call myself the beacon chain Health consultant it's a made-up name doesn't matter um and I'm here with my good friend Ben Edgington hey Ben how are you hey face uh for those who don't know me I am product owner of techu the uh ethereum client at consensus and I'm also proprietor and producer of what's new in eth2 which is something I've been writing every two weeks or the last almost four years now unbelievably um it's it's a great newsletter every Fortnight he produces the newsletter and

then about three days after every Fortnight that would be coded as fortnite plus three um we uh did this kind of walk through where we chat about it um because it's it's so rich in information um and uh Ben is able to give just such great context to it and it also gives me an opportunity to add my little decentralizer spin to it uh which uh I think it works for everybody uh but we were just talking about a poab um I thought we were talking about chocolates sir yes I mean pardon me we were talking about Toblerone chocolate so Toblerone UK uh had this promotional

where they're releasing some kind of new Toblerone chocolate brand um and uh uh I guess I someone noticed on their website that they're offering a PO app uh but they're only shipping to UK and I am not in the Kingdom uh and so I had previously sent Ben something many months ago uh and so I still had his address on file and I was like Ben's the only person I know right off hand in in UK and I have his address so I ordered um Toblerone chocolate bars to be sent to him so hopefully I can snipe one of those uh po episodes what's the deal

here I I get to eat the chocolate you get the perhaps is that it sure why so I'm gonna need a write-up like a post-mortem about the chocolate though right like I think I ordered three of them because I oh there are four in your family aren't there uh yeah we're not all big chocolate eaters so I'm not even convinced that it's chocolate I think it's caramel oh that's right it's um some kind of gold thing yes that's a promotion yeah yeah I didn't look too closely me either when it comes to poet I uh I throat inhibition out the window and do

whatever I can take into that one yeah absolutely yeah uh anyway that's a bunch of fun hey but I did get a tracking number and I think they're going to be there in a day or two so uh excellent good for you I look forward to it uh oh and the newsletter we talk about the newsletter too yeah yeah we better have a look Ben and I were uh greeting this morning and he uh well this morning for me and he started bringing that up and I'm like wait stop let's record it the posterity I'm gonna share my screen hopefully the right one oh oh I'm sorry I'm in a sub tab here that's why I'm confused all right pardon me while I

okay does that look like the right newsletter I can say something Edition 98 yes indeed great um and so essentially what we do is um Ben spends hours writing it I spend about three minutes zipping through it and uh we just comment on it but the most exciting thing ever uh I love this line right here why did I highlight this one we are all systems go for girly Prater um yeah yeah it's happening um it's we're in the sort of the Calm before the storm right now but uh targeting uh 11th stroke 12th it seems to be coming a little earlier uh on that

chart so so the girly TTD so it's it's not a proof of work Network it's a proof of authority network but it still has a difficulty and the difficulty of each block is either one or two depending on whether it's produced in order or out of order which is um proof of authority thing and and that's sort of Random so we can add up these ones and twos and get a sort of terminal total difficulty that we're heading for um but it's still a little bit uncertain so there is a there is a margin of error around when we might hit it so currently looking like Thursday yeah late on Thursday my time

uh is that I guess that's UTC not not adjusted for your time zone right so uh let me see remember yeah let me see it's the same here it's not adjusting right yeah that'll be UTC or GMT so uh yep yep uh should be at a decent time of day for you that would be um quartered one in the morning for me oh wow and there's not a lot of flexibility with this because they they can't really add but the only thing they can do is is turn off uh yeah without uh miners they can't actually add more right so yeah and that might not even be predictable I I have no idea what effect that would have on the TTD so on the on the difficulty so yeah it is what it is

we can't speed it up we'll slow it down we just uh have to sit tight um yeah within at least it's pretty diesels somewhat predictable yep yep um yeah and then within the next 10 days or so the uh that girly Network merge work will happen and basically that's it that's the go no go for for the real thing right um and the thing people should know is uh Gurley was very much decided to be the easiest of the merge test Nets they did hardest to easiest um which I obviously I don't want to um I don't really believe in jinxing but I don't want to jinx and say you know it's the easy one and then it blows up in our face but ideally it should be the easiest merge right

yeah um yeah not sure about that the so interesting factors um we we've sort of postponed it as late as we can whilst keeping a Target mainnet um dated mind because we wanted to have basically production ready code for both the execution and consensus clients so the the um yeah we we could probably have done the girly merge you know two three four weeks ago but the idea is just to push it back so that we can all have the most um closest to production ready client uh we're merging it with the Prada Network which is what you're looking at here which um has been running for quite a long time so this is um

uh and a lot of the validators on Prada Network are controlled by the client teams um so we run 20 000 I think for techu um maybe more uh and the other client teams run you know huge numbers of of validators but there are also a lot of you know individual stakers and other people using Prada it is the beacon chain testnet right so anyone who's testing staking infrastructure on the beacon chain is using prior to so there's a lot of different organizations involved so on on that basis it could be interesting to see if everybody upgrades in time configures in time gets everything gets everything right it's a lot bigger I mean this

um 350 000 validators is a lot bigger than supposedly had um we we put up specific new Beacon chain networks for the other two for Superior and Roxton um and and this this dwarfs uh those ones so yeah it's going to be different that you know the thing that that I think about is the number of discarded the number to start discarded crater Keys uh like I know that I myself I've spun up several in just whatever like maybe I've I've nuked that virtual machine and been like where are those guys at um or you know just something silly and yeah I I guess that's that's not really a terrible risk uh but it probably does

account for this 10 of yeah missing validators and we can expect that to fall again maybe by another 10 um after the merge for people who are maybe they're running cloud or maybe you know they've got something in their closet that they just never thought to upgrade yeah just not that interested yet yeah uh it definitely gives a bit less margin for error than than we've previously had so for reference the the mainnet beacon chain runs at 99.5 plus participation um pretty much all the time now um yeah it would be nice if people uh exited their validators properly but uh there we go 99.7 it's unusual maybe it's just not falling I think it's still

catching up um yeah yeah so yeah that that gives you know we could we can lose 33 of the network and still be in good shape um on on Main Network as we can only only lose you know like 23 of network on prata and still be in good shape so um there's less less margin for error but uh yeah interesting times um no no real predictions about how it would go yeah I mean it'll be fun like I and I think you won't predict I'll predict that the clients will operate smoothly but the validators the the number of upgraded updated validators will be

lower than expected and so um I I look forward to something like 70 participation with Finance with finalization so uh okay we should do the sweepstakes uh on this the uh yeah the validators will perform flawlessly but the people may not um it turns out I I put a lot of faith in my own judgment and it's always wrong so I'm a bad Gambler uh let's see so prior to the merge on uh the girly prata merge we have the um upgrade of the beacon chain so the Prada chain to uh with What's called the

Bellatrix upgrade so that just prepares the beacon chain to be able to absorb the um execution block the execution payload as part of its data structures and um it um changes a few constants so the currently the penalty for being slashed the initial penalty is like a half an eighth where that goes up to one eighth and and a few other bits and pieces but it's mainly just getting the the the data structures ready for um post merge um merged Beacon chain execution chain it's really important that people recognize um Prater is doing Bellatrix now um Gurley will do Paris soon uh but then later mainnet will also execute

um Bellatrix and Paris so we're going to see those uh Bellatrix in Paris in two iterations first on Gurley uh Gurley and Prater and then on Main net and Beacon chain so yeah yeah the the Paris thing on uh the execution chain is not actually a hard Fork um it's so clients will need to be updated and they need to be running the Paris compatible code but it doesn't need to be coordinated because it doesn't become active until the merge event itself whereas on the beacon chain We are Becoming active with the Bellatrix code ahead of time um which is yeah we've got two uh gold all the code that manages the TTD and stuff and the fork Choice

um uh when the merch happens that needs to be ready cool um so yeah we're looking at that probably between August 6th and 12th um I'm hearing August 10th and I guess we just looked at the uh yeah the the bordell uh it's definitely towards the end of that that window yeah um Bank listen e-staker are working together to do a community call um I don't know a lot of details about it we we are working with um bankless but uh we're very much letting them take the lead uh and so um as soon as they publish more details from their side then we'll definitely get them out there um but I'm I'll definitely be there it'll be the third of a fourth the third

of four uh uh what does it trail to the merge update uh po apps that's that's really like I'm just here for the co-apps yeah but we did a poet for supolia one four robs or robsten sapolia um we'll do one for Gurley and then one for mainnet indeed and they mentioned uh bankless spoke about this on last week's roll-up so uh they know it's coming yeah yeah I think you know it's uh much of this is just it's been done like we we've got the flow of it so um we all kind of know what it's gonna look like we all kind of know what it's going to happen uh when it's going to happen

um we just haven't like uh etched it in stone but it's it's there it'll be there so song a day man is teed up and ready to go yes he's very stoked about it um so I'm excited and well so speaking of that uh for mainnet our plan is to do some of the old fun stuff um well the mainnet plan is uh the daily way bankless and each Staker streaming from the ethereum foundation stream so we can all be together in one place and that means we can have all of the guests in one place and they're not you know hopping calls um we're planning to do a um a poet the art canvas that we can all kind of so you know everybody

contributes pixels and then we have one magnificent picture in the end um and uh yeah and Jonathan Mann on on the final call we think he's he's going to do a a planned out song hopefully one that has had you know a lot of Polish to it um he often does impromptu songs which the audience loves but they don't always have the same uh musical qualities so I was like if we're going to do this on the EF Channel we cannot uh make the whole song about trashing other chains or like you know foul language or whatever so um anyway that's good yeah there's going to be a lot of sweaty Palm to devs on

that call I can tell you uh already a lot of a lot of uh devs who are uh going I was going to change my audio to the the placeholder like you know when devs are going to placeholder screen you know that you know they're they're starting to worry about something yes Shadow forks and you know everybody's been saying Shadow Forks are old hat but now I feel like they really are yeah they definitely routine now um you know we're still finding one or two little issues they're definitely still worth doing um and you know just learning about the configuration and and so on uh slightly weird

uh Edge case where nethermine showed up in the Last mainnet Shadow Forks number 10 um you know not not a catastrophic issue um but a temporary deadlock um was discovered and you know basically the spec wasn't clear enough on some Edge case and uh that's a uh that's easy to clear up so uh um but yeah routine and it should be at this stage if we were still having major issues with shadowfork said you know we we wouldn't be looking at emerging potentially you know six weeks or something I mean it's uh it's how it should be I'm not sure whose comment it was it may have been you we may have talked about this last time uh I want to say that I heard it on a talk though uh the idea that we should have been doing

Shadow forks for like the past seven years and it's just this great idea that now makes so much sense that will uh really improve our progress going forward it's um a brilliant idea uh I think Marius may have invented it from the Geth team um it was uh uh yeah just stroke of Genius and it's been it's been really helpful yeah great [Music] prepping for the merge so exciting this is a lot of work not I don't want to say it is a lot of work it is some uh required work for operators kind of critical work that they get it done not that it's hard but things that

really need to get done yep uh and this girly proud to merge it's your last chance is going to be the closest thing architecturally to the real thing and um you know people would be well advised to uh spin up a note and give it a give it a whirl either you know as a Staker or just as a non-staker uh setting up your consensus clients and your execution clients so they can follow along and make sure that they can continue to progress through through the merge um there are a whole bunch of just fiddly Little configuration things to get to get rights to to make it all happen so yep um and guides are starting to appear as better and better materials on on things

and clients options and client documentations so just settling down now so it's all is we we're getting to a good place so we're hosting a lot of support on uh the Gurley Prada channel on the e-stick or Discord um I don't hang out there much but I know they're a really a whole lot of bright Minds there um it's so great to see active support um Chris hopcroft put his uh guide out do we know who he works for like I see him a lot but I don't know what his affiliation is nope nope uh I met him very briefly in passing at uh def connect but we didn't stop and chat so uh that's uh that's all I have yeah he

runs techusa well obviously a good feather yeah of course um let's see uh we did we by me I mean by we I mean uh e-staker who I really mean is uh Remy Roy and uh York Downer or thorsten uh and along with uh onvetica did a merge valid theater prep Workshop my my favorite thing about that go ahead yeah I just tuned into it sort of while I was writing so it's on in the background so I wasn't following uh Blow by blow but I could hear that there was some good stuff uh going on some good questions and covering a lot of grounds so uh yeah that was uh that was great to just uh uh hear that and see the

participation there are a couple hundred people sort of on on the channel as uh as it was happening and I guess a lot more we'll we'll catch up later yeah as kind of an organizer for each Staker I love to see subgroups Branch off and just do stuff um and so like they didn't even invite me they just sort of said hey you know this is a thing we're doing and and so that that just gives me so much satisfaction for the East taker group that they are uh supporting without sort of requiring everyone to be on board for everything so at some point the time the day will come when they don't need you any more fears and then you can retire to a beach somewhere I I think that happened about eight months ago inferior is doing weekly merge workshops

yeah these are these are good I mean a huge numbers of people um register I don't know how many are Bots I mean to be honest but they're like 10 000 people um were present um uh one I participated in recently and it's uh yeah it's kind of hard to believe that they're all real people but there's a lot of activity in the Q a and in the um in the chat that appears real so yeah um definitely popular and good stuff um they're really getting to the it's not not your superficial you know bouncing off the surface of its stuff they're they're taking it slowly enough to they can dive into specific topics with people who really know what

they're talking about uh week by week so yeah recordings are available uh for the past ones you can see the list if you click on the list it will take you to the agenda page you know where the the list of uh the whole list of talks is um 10K participation is nothing to uh that's pretty impressive what do you think drives that um so the next one is tomorrow execution client and consensus client um and then uh there's um a later one I think the following week is aimed at the larger operators like uh staking Services you know running multiple clients running

multiple validators and so on so cool Carrie who's one of my favorite people lately um sharing documentation about the merge and these I think are pretty standard but great to see reiterated indeedy yeah lots of information coming together so uh uh yeah go out and find it uh so sea monkey is one of my favorite well I to me he's a redditor but he's my favorite redditor because on the the East Finance daily um he gives these reports of like you know

um you know these These are the combinations I tried today these are these are the successes and failures so I'm I'm just wildly appreciative of what sea monkey does uh yeah this is insane uh this is pointing to 20 combinations of clients uh consensus and execution clients uh I think there's one validator of each so it links to the beacon chain um page for each uh each of them or yeah I don't know this is 20. so oh they're 20 pending so they've just been deposited a year ago yeah that looks like dashboard for all of them um yeah presumably they've been running on prata um there's something about that doesn't

look quite right yeah I think the uh the deposit is still being processed or something but uh here's something more more new yeah not not active yet I think so uh yeah deposits still being processed I think just get through it in time so amazing to see like uh I I don't want to say anyone can do it but anyone with the commitment can run those combinations and it's it's really impressive um yeah so this is a config script that just lists out the command line stuff for you know the options for all of the client combinations and so useful cut and paste stuff if people are uh um wanting to uh you know just somewhere

easy to grab the various options for various clients this is the real gold too like anybody can set it up but then to actually share your documentation about how to do it quickly that's uh very valuable so thanks again to see monkey rocket pools guide oh yeah yeah it's a good reminder everybody's got to take some action um including rocketpool stakers I I have to confess if I wasn't already obsessed with rocketpool I would get obsessed with rocketball like it is um in terms of running clients of any choice it is so simple to select a client it installed it all for you the background in Docker and then if

you want to switch clients it has like this uh a terminal user interface it's you know like an ANSI screen where you just select and you're like I want to go from Bay suit to or I want to go from guest to basic or I want to go from Nimbus to techu and you just select it and it does it all in the background um I I've just obsessed with the efficiency of their operational indeed yeah there's a bit more rocket pool in the in the next section as well so they've been uh busy I'm just getting warmed up yeah I'm just kidding uh Nexus Mutual uh and this is a steak wise kind of thing or actually it's not stickwise it is uh I guess any any provider was less than

10 or something uh yeah I didn't look uh in detail of the terms and conditions I did have a quick look through um and yeah I mean it's kind of interesting um slashing uh slashing Insurance um I know there are conversations going on you know around the place and I guess it's a way of de-risking um I don't know what the costs uh well actually they do talk about the costs I didn't do an analysis of it but uh uh yeah uh if you are concerned about slashing insurance it's probably worth a look stake wise certainly thinks it is yeah great uh and I I wonder I guess I get the feeling that's more about instilling confidence uh and that's kind of what insurance is in these cases it's

not really that you expect to cash in it so that you you instill confidence in your user base that you are covered uh for something you don't believe will ever happen yeah it's risk management you're subcontracting your risk management to an external organization which is uh which is fine I mean you generally on average you pay a price for that but uh you know it's generally better to self-insure if you can afford it but um uh on that basis you know it's just a normal risk risk management activity which uh um suits some people yep um I mentioned rocket pool as if it's not easy enough fast contract has gone through step by step everything to set up

a mini pool or a node in a mini pool is using his user personas his parents which is which is correct which I love yeah yeah um not sure what my my mum would make of that but I saw a much earlier version of this but he is really uh filled it out that is really great um and I love that they're using 2fa um login so I've I've set all of my everything in my home up now is is using uh well I should not reveal my security apparatus but I'm using 2fa on everything now I just feel so much I feel I believe it's so much more secure than uh good old-fashioned key sharing

very cool that's again I really love this guide uh if it's not easy enough it's even easier now the three-fold scaling plan we don't have to read this but you want to hop in this real fast yeah this is really good um there was a podcast which gets a um uh gets a nod in at the end of the thread which I heard a few weeks ago which was um I think it's Ryan berkman's who was talking about I quoted um his take on Lido which was very negative but he also outlined quite a lot of issues that rocketpool had you know what because the question is if if we're down on Lido then what's our alternative what you know what can we

offer as a as a as a solution to that and he had some suggestions for rocketpool about um uh how they might consider uh improving their um you know lowering their Capital costs and so on and just you know attracting a big user base and uh uh and rocketpool have come up with uh with a brilliant really really strong response um so so one of the issues is that um if you run a mini pool so you're running a client a rocketball client that others will will stake um ethon you need as the node operator to put down a 16 eth Bond now big basically because that's a maximum your

node can ever lose uh absence slashing so if you neglect it completely and the Network's not finalizing then it will run down to 16 Ethan you'll get kicked out and then you'll you lose 16 Ethan and your your your others um uh can be made whole uh you know the ones you're hosting from that but um they did a sort of but this is a big drag right you know if I've got 16 Heath then I can only run one mini pool but if we can lower that requirement that that Capital requirements if it's only 40 per minute but now I can run four which actually quadruples the capacity of the rocket pool Network for the same amount of down payment which is very cool so

they've done a sort of risk management exercise and looked at actually is it what is the likelihood that actually anybody will ever lose 16 each just by neglecting their node and basically is negligible so um they've done a sort of risk management exercise and and calculated the uh an optimal amount which they've come up with four and you know absent protocol changes that seems like a very reasonable approach so that that should unlock my understanding at rocketball is it's not demand side which is a problem there are plenty of people who want to stake with rocket pool but it's supply side there just aren't enough node operators who are willing to put down the uh the required uh ether

so that should unlock a a lot more network capacity so that's number one is that a fair take is that your oh yeah um I I the only the only thing I would say slightly differently is uh the economics do tend to shift back and forth between um supply side and demand side so rather than what you would you know eventually find as an equilibrium what what it seems to happen to me in any way is is there they tend to stair step um and so that that does cause hurdles and there's when we when you zip through these I also want to bring something up about um my idea of the or my my questions about what's at stake uh so I'll let you

do that really quick and then I'll ask my question yeah so um number two a new staking as a service architecture I was a little hazy on this I didn't take the time to you know dive into all the details but it's sort of abstracting um the the rocket pool architecture to make it more more flexible uh in terms of um what node operators can stake uh whether it's RPL or eth and and you know what what they're prepared to be long in as it were and just sort of making the um the hosting of nodes more abstract um which which is interesting I I haven't got a lot more to say on that maybe you can help me out but the the

last one is um very cool which is uh I so I run my own validators uh here and I would quite I'd be quite interested in in running a rocketball node but once you've sort of staked for your own validators and you can't currently unstake them and get the get the state back because we don't have that in protocol so I've got no way to spin up a rocket pool node without actually starting from scratch you know um I can't transfer my current validators into it so very simple um process a for anyone who's got the old credential type the a type zero credentials which is where you needed a BLS withdrawal key and so on

there will be a facility in protocol to migrate that to a normal ethereum address and you can just point that at the rocketball contract and you can transfer your old validators into rocketpool which should you know give the opportunity for um a few hundred thousand um uh stakers uh if they wish to to move into rocketpool because yeah we should recap the history I mean rocketball was delayed by a year right um so they didn't become active till the year post merge uh post Beacon chain start um so you know there are a lot of people who might have wanted to state with them but but just couldn't now they can now they can well not now but when this is available they will be able to transfer

in well one of the most pressing criticisms of Rocket pool is that they are so slow and deliberate um which is beneficial if everyone is slow and deliberate but if uh one team is slow and deliberate While others are racing to the Finish Line uh then you know it's an imbalance um but so I've I've had this question that it's been swirling in my mind and it ends up being related to the uh the nothing at stake problem and that it's kind of like as a as a LSD participant um and I'll pick on Rocket pull because I am you know rocket pool I'm very close with him so I'm I'm happy to to poke

holes in them um but it really does apply I believe to all lsds as as a as a liquidity token provider uh it seems like it sort of nullifies the concept of having something at stake in the protocol um because as a liquidity token provider I can dump anytime that I want and so those coins are already deposited in the contract and let's say that there is while admittedly unlikely a slashing event against those staked coins then it seems like the person who pays the penalty um well the person who gets away scot-free is the person who dumps

liquidity tokens as soon as it's noticed like maybe I have a bot watching it sees a slash it dumps those liquidity tokens I get away scot-free and then the people who are left holding liquidity tokens as in Grandma and Grandpa are the people who are most likely to suffer the benefits or the the loss associated with the slashing is that does that make sense or am I missing something yeah uh absolutely right there is this sort of abstraction away of the um uh of the risk is the risk is moved around it has this kind of like 2008 uh feel about it right you know where where is the risk in the system it's a show

game right uh yeah yeah um uh and yeah absolutely right there was a good article written along before maybe two years ago long before the beacon chain was up and running I think it's called the death of Ethan it was a d-e-th of eth um which predicted this effectively the the hazards of liquid staking derivatives um and yeah uh you're right and you you have this I can't remember what the um the the moral hazard thing I mean it's sort of abstracted away the people who are actually um taking care of the risks and not the people who suffer the consequences for the risks you know and uh it is a problem but I I think

there's almost nothing you can do about it I mean apart from being aware of the risk the resurfacing the wrists and so that people can price it in correctly um there there is almost nothing you you can actually do about it in any sufficiently you know in any protocol that's that that has functionality like ethereum yeah um you can't you you you can't tie that risk down and just keep it in one place it's always going to diffuse through the system yeah the only solution I came up with it's it's very much a scorched Earth solution um and I you're not in favor of sports

and Earth Solutions but it is literally preventing the the implementation of uh liquid staking which then requires people to uh hold that Bond but then you know what they'll do they'll just go to a different chain and uh migrate the risk to that chain and and still allow for liquid staking on each so I guess you're right like yeah I mean by Design ethereum basically cannot stop you doing anything I mean it's Turing complete I mean effectively you know there there will be a way you can uh you know if a token has certain properties that you don't like you just wrap the token I mean it's yeah

yeah it's difficult to get away from I think maybe the the and I know we're getting a little off topic here and I hope that's okay I I think maybe uh the best thing that can happen are some minor hiccups now kind of like the anchor and this and the state.us getting slashed like hiccups now that educate people and help them realize the risks but well it just makes the protocol like weak yeah you think so wouldn't you I mean so one of the big factors in bringing down three arrows Capital which was you know a seismic event in this ecosystem right um was staked drifting from its not Peg but nominal Peg

um to to eat right because they they leveraged up on the fact that state eth was um you know should trade and converge with with Ethan the long term but the market can remain irrational longer than you can remain solvent right so um as it's as the staked price drifted from um The Ether price due to foreselling by people who are getting liquidated um they they got liquidated and uh that was very very expensive so massive risk event due to um not properly pricing the cost of liquids taking derivatives um did anyone learn any lessons of course they didn't I mean that's the

interesting thing about that that's an external forces incident that car caused the de-pegging now imagine an internal forces incident that causes a de-pegging and I think that that is something with more potential to cause so what we saw was a small spiral um I think an internal forces incident I guess what I mean is by that kind of Provider being slashed um will will cause uh is likely to cause more significant Panic uh but I guess only time will tell right we'll have to see how it works out yeah fast is fascinating we're building a whole new Financial system it'd be nice not to repeat the mistakes of the

of the formal one but I feel I feel we will again and again you know I think it's easy to overestimate what the core protocol is capable of um you you know you want the protocol to to be developed to prevent these things uh but it we just we're not at that level of sophistication yet um it just doesn't exist so it is what it is sorry all right Alex Stokes Alex yeah good old Alex um he's thinking about the right problems so uh yeah this is a kind of issue we've had on the back burner for a while

um weak subjectivity is a thing in proof of stake so you have to trust if you're bringing a new node online you need some people you trust to some degree uh to to be confident that you are joining the the right Branch the Right Fork of the network in principle I mean in practice I don't think it's going to be a big deal but in principle you do how do you build an um how do you automate this this trust and Alex is thinking about that so uh yeah very cool um I mean that article on weak subjectivity is vitalik about 2014 I think writing about it it's not a new problem um but we just putting down practical

solutions to it it's it's interesting I guess the the deeper I get into this it's interesting to realize how much of trustlessness is bootstrapped in Trust um like yeah I get the need to get away from it but it is kind of our bootstrap yeah uh every time I transact with a D5 protocol I I remind myself I have not audited their code exactly yeah I basically have no idea what's going on at the back end it's supposed to be trustless but I actually I'm putting a lot of trust in it um but it is in principle and trustless right it is transformative somebody can audit it which is not which is not what we have today I mean the financial institutions

are totally opaque um I don't talk about this a lot but I I interacted with a malicious contract and had had a pretty significant con consequence and it yeah it's it's like it may be trustless but I did trust it and it abused my trust and so we're not quite that not quite as far away from trust as we'd like to be but I guess we're getting there uh my poapps that's what I what I lost but they all came back it's okay ah very important everyone does know you need to run a

execution and consensus client post merge but tell us the exciting but here yeah I think we touched on this uh last time perhaps because the guest team had been working on something a bit similar but now the Nimbus team who are really championing Lifeline technology which is great along with the lodestar team um yeah in future uh if you just want to interact with the execution chain so you want to um a host a smart contract or application or you just want to get you know know what your account balance is or whatever it is you don't you know under today's well under the merge architecture you have to run a full consensus client and a full execution client and the consensus giant client

checks that you're on the right chain and and that everything is is correct and the execution client you know checks that the transactions and you know maintains the state of of everybody's account balances and so on in future you may be able to run instead of a full Beacon though just a light client to the beacon chain which is very low data requirement and very low CPU requirements so one of the lessons learned from ethereum one is that the light light clients are really hard to build in ethereum one it just wasn't designed with like clients in mind ethereum 2 from the ground up and yeah in in the stuff I've been writing in the book

um I call this out quite a lot because it's surprising where it pops up ethereum 2 has really been designed to be like client friendly so every data structure we have every cryptographic algorithm we have is is designed to make um to make like clients light so you can have this you know potentially in browser or you know very small piece of code potentially built into an ethereum execution engine which will consult to the beacon chain and it can validate based on some minor trust assumptions it can just get a small amount of data from the network and validate that you're on the right part of the the vegan chain and your um your execution client is is following

the correct State and has all the correct uh information it needs so that is pretty cool um it feels like in a couple years when we have statelessness and like clients that the dreams of running a validator on something like a toaster now if you trust your toaster you know with 10 million dollars worth of eat at that point that's great but um the the computing power required the storage you know everyone's panicking or everyone's worried you know well I only have a one terabyte drive and I really do suggest a two terabyte drive for home stakers now but that there's going to be a cap to that it's not going to grow forever

yeah if we deliver on the roadmap yep now we'll deliver now it may take 40 years right but it'll happen um I I always love Mega Labs such great contributions yeah yeah um that'd be doing yeah just testing what you're talking about with Rocky pool earlier about uh switching clients uh very easily they've been doing more thorough sort of analysis of it so uh look forward to seeing the final results uh of that but um looks good I mean so you know as client teams we are um we initially started going our separate ways implementing very different interfaces and slashing protection and everything but one of the things that I think we've done well is

cooperated on things like common slashing protection interchange format so you can export the slashing data from Prism and you can import it into teku and um just putting these uh mechanisms in place which are user friendly um has been a very big part of our collaboration and yeah it's great sure and help me please correct me if I misunderstand this but you know that I generally consider myself a homesteaker not an Enterprise Decker and so um this article it talks you know about migrating 250 validators and how you would want to migrate the

um slashing protection database as a home Staker I generally just shut my node down for you know 12 13 minutes and then reboot you know get get back online uh and those are kind of different requirements because you know as an Enterprise taker if I'm if I'm coinbase or if I'm cracking and I'm cracking I'm running thousands of validators I can't afford for my customers to see me offline but if I'm you know Fizz at home and I have you know one validator just shut it off for a while and restart it is that kind of a fair assessment yeah uh that works the two Epoch kind of uh shutdown should be fine as long as you know your system clock on your new box

is okay and the slashing protection will save you if you you know if you change the time zone between your boxes or something um so it's worth going through the smaller extra step if you are migrating between boxes just for complete peace of mind I would I would recommend it doing it but uh um I I should probably recommend that it's best practice instead of doing it fast and loose but it's always that balance you know you there's always you can always take an extra step you can always you know there you can expand everything you know you can build your client from Source every time

but you end up making trade-offs and the trade-offs that you make you know you can generate your keys on a completely air-gapped machine in a cave that's you know shielded from RF interference but you know people just generate keys I'm always trying to find that balance of what what is best prac or what is good practice versus like what is the gold standard yeah yeah I mean and 999 times out of a thousand you'll be fine but uh just occasionally um somebody will you know have a Trojan on your PC or something and and pick up your back to risk mitigation right

um opal has launched their Athena test net um I saw this but I didn't hop into it um what do you think uh yeah great to see progress so this uh distributed validator technology is coming um so SSV Network and mobile are both working on it we're moving towards sort of standardizing the behaviors um this I think will be uh yeah having options is great so this is an infrastructure option so you can run a resilient reliable validator um made up of multiple nodes and you can lose one or multiple of those those and still continue functioning so no

downtime um so that this is um it's a great technology and uh yeah good to see progress I don't know what else to say on that one yeah I'm excited about it too I I'm probably less uh less obsessed with DVT Technologies but I do recognize that uh they are part of our future so yeah I I uh I'm not excited about them but I do welcome them it's kind of a hard balance yeah yeah I'm kind of saying um and it's I think what I like about about these guys is that they they have a multi-year future Vision so I'm often in the what do the next two to three months look like mode you know

kind of what features are we working on and what what what's the next thing for techu and you know relatively short term thing but this infrastructure layer is a multi-year play what what will staking on ethereum look like in 10 years time uh what are The Primitives we need what are the architectural options we need uh how can we make it truly resilient robust I I just figures My Hope for it in 10 years my dream is that more people will bring these validators into their homes um run them you know independent validators at their their office or their company

and you know the the feature of something like DBT is to make it more Enterprise grade and I think that's where that difference is like I know there will be Enterprise operators but I really want I want pull that back into homes but yeah I mean if you're if your toaster is going to be a validator right it's better to network it with a you know a bunch of other um validators around the home so isn't it you see a toaster around your refrigerator and you and your television There Are all uh you're not gonna get robbed right oh my toaster or or your toes were in my toaster you know both yeah being validators right in in the pool so that

if you know um it's on Low End Hardware if it goes down or I need to reboot my toaster or you know whatever my internet goes down but I'll validate our Collective validate keeps running so that's the promise of distributed validated technology is that it gives this resilience even if individual nodes are unreliable which um it might be really good for homestaking I mean it could open a whole new kind of style I really imagine that as the promise versus the delivery um you know the the promise is at my home my validator my toaster and my refrigerator are linked but the delivery is more likely something like a t

and uh Ford and Tesla are running you know all of these validators and kind of taking it to you know 78 levels above people doing it at home which is you know the difference between what we hoped would happen versus the actual implementation so um yeah I guess only time will tell we will see we'll see nothing is is more I need to find this clip from I believe it was vitalik at ECC um where he's saying you know you'll be able to stake from your laptop and if you're traveling it's okay like you might be offline for a third of the time but it's no big deal that has always been my like dream of staking and at

some point I have to acknowledge that reality will be a little bit different yeah yeah but if we had uh well we're probably going a bit of a tangent here but if we had uh like eight people's laptops in many pools in different time zones uh you know in a in a distributed validator technology cool then you know when you're offline and but I'm online and enough people as long as enough people in that group are online then we're good so that would be absolutely sexy to me like I that's yeah that's a great uh yeah a great optimistic perspective that's crack on I'm on these tangents today I hope it's okay with you I'm enjoying it it's good with me I just I hope our audience have got the patience

I don't even know if we have an audience though I know you look I know you look no no not anymore not anymore yeah we'll check and see hey uh so uh I've never said this stuff before uh comment if you if you don't mind the tangents or if you do like the tangents and we can tighten it up one way or the other um 48 44 and done in that yeah Paulina Paulina has become a bit of a troll these days started off with some very um uh incisive uh analysis of uh layer 2 stuff but has becomes kind of self-confessed uh asking the hard questions in some

ways yeah which is interesting so this is um this is basically a suggestion that we ditch dank sharding um from the robot or push it out 20 or 30 years um and that actually we just do proto-dank sharding 4844 which is an in protocol uh which is basically a block size increase um or a temporary block size increase because you can throw away the extra data after a month and paulinar argues um this is going to be perfectly adequate for the foreseeable future it will provide plenty of data space for um Roll-Ups um in any conceivable usage patterns uh for

um the coming years which is interesting I mean as I say here it's a kind of echo of phase one and done which Casey put out there like um three or more years ago um and it ended up in what the the route we went down um and uh because we had this phase zero phase one phase two road map yeah um and and Casey's argument was ditch phase two which was kind of in protocol um execution shards and and that's what we did we pushed that to Roll-Ups and said okay that's that's that now outside protocol um and um that's what happened so yeah we'll we'll see

um we'll see we'll we proceed a step at a time we deliver we reevaluate we re-orientate pivot if necessary so it's the ethereum way I I'm mostly amused to see that the phased approach is no longer in my Consciousness like it used to be all we talked about and now it's it's totally totally deprecated yeah the uh warning not to be too wed to a particular roadmap uh explainers ethereum pools does these really great is it an image yes thoughts yeah more good stuff I mean just looking

at the analysis of how um uh people are doing in practice on the beacon change so um yeah if you're interested in rewards and penalties take a look uh yeah I do like ethereum Paul's stuff uh there's some good stuff coming out of there for me it's a struggle uh because it's the focus on rewards um and the focus on uptime and I I'm not focused on up time and rewards I'm focused on Network Health and it's that whole idea that uh um the most robust Network we're going to have is not one with um perfect rewards that's that's a

Legacy Traditional centralized system the most robust Network we're going to have is going to be one with variable rewards people offline um you know people running validators in places where hurricanes and tornadoes happen in the real world so I I do struggle with yeah I kind of get you but the rewards are there for a purpose I mean they are to encourage you to yeah a degree of it but yeah I mean I agree about um definitely agree about kind of arguing about the last you know decimal point of of uptight but um and people do get obsessed with that which I think is crazy but in general

the the network rewards you for being up for a reason I mean this is you know um XI Jen so I do follow Fijian on Twitter um her FAQs are great yeah this is good good Insight it's all all accurate um love it yeah you changed your pfp and now whenever someone changes their pfp I don't know them on Twitter anymore like there should be a law against that um crypto Texan tackles eight misconceptions about the merge that was a great yeah just lifted from the uh ethereum website but uh in in uh buy size for uh again that's what people need and stuff yeah um all right and

ACC I didn't catch a lot of these I did watch metallics um they're a bunch I want to catch up on yeah what did you like well I mean here here are the ones that I uh that I caught up on and um enjoyed them all and um uh yeah uh absolutely nobody I bet nobody has clicked on any of these links because absolutely nobody's complained about me Rick Rolling them so did you Rick Roll it so as I'm starting to wonder who which uh well it's the bottom one the Rick maybe it's two signal maybe it's uh so it's too obviously I have a confession I won't no I won't confess it because we are too

um ah we're too public I'll tell you afterwards I've already baited I actually thought that was someone else and I checked to see if they had actually done a talk and they did not so I was like oh that's weird um anyway we should move on um cat herders always producing great content yeah yep yep yep know your client this is great I mean they just keep keep on chugging with the with the content um Adrian recorded the techu session last week should be up uh any day now so uh uh yeah terrific okay I I have a challenge for you I can't say it out loud because it will

so next week I'm gonna suggest this afterwards and it'll be a fun challenge for next week uh yeah I because I can't spoil it it will ruin the surprise for next week and next fortnight so regular calls we yeah we know about these all quartets consensus call breakout um yeah lots of calls going on I can't keep up with everything I mean the major things I just kind of note and put links to agendas and things um uh I attend most of them but not all of them um so just yeah and I guess what we really need are people to do really good Twitter summaries like these bite-sized chunks that that really is how we get information out lately Yeah Tim Baker is great at doing uh most of the the big

ones the um the all cool devs for example and Christine does some pretty good summaries as well though she's taken to more long-form stuff now and I I also look forward to uh Perry's uh follow-ups there or Perry summaries they're they're very valuable of like the the shadow Forks yeah the reason I include all this stuff is is I have this kind of fantasy that you know in in 100 years time there'll be a sort of digital archaeologist who wants to kind of um you know track the uh origins of the the ethereum story and we'll be digging down into the archives and we just will want to find where all the all the materials are and I I sort of fantasize that you know what's new in eth2 will be

the place that kind of links to absolutely everything that uh that's required in this uh um this digital archaeologist will be profoundly grateful to me for for having put it all together all because of Ben edlington it's a valid fantasy I mean you know it's a public good you're doing a future Public Service um it is it's important to line those things up um because uh getting the time that different things happened like in the same scope is very difficult yeah even at a couple of years distance is kind of hard to remember how things played out right definitely uh stateful

works um yeah a good good old Trend uh has been uh pushing on so I don't know much about this apart from what's written written in the article but uh um a series of nfts uh and all for um public goods well ethereum public goods which is basically the the protocol Guild which is funding um a hundred and well it has been 110 which we just onboarded a few new members but it will be maybe 120 core devs of of ethereum um which has been a great initiative so uh yeah um Tim and Trent have been very busy

um bringing in funding streams to the protocol devs a protocol Guild uh and yeah have some very creative activities this is uh this is one of them uh so I I absolutely adore Trent um but when I read these things sometimes I'm like I just want the link to engage like there's a lot of context and I'm like I where's the buy button like just show me the buy button but he buries it like it's in there somewhere you just have to find it if you're listening [Laughter] um I think we made it to the end I highly favor yeah so um Trent protocol Guild uh though that

project is obviously very near and dear to our heart uh so anything you can do to support them also supports the ethereum ecosystem so very awesome um we we come to to time but since we've talked about perhaps quite a lot let me just um she'll uh left Harris uh yeah yeah if you want if you want to um a very exclusive pull up uh I retweeted lefteris's um marathon thing he's running the Berlin Marathon which uh in in September uh and he's fundraising for a great cause which is for infant um heart uh yeah babies who have um heart defects and um

there are perhaps that's the point the point is that he is offering perhaps to uh those who are donating he's harnessing the power of the poet um so he's aiming to raise like 42 000 um die I think or you know dollars worth which is like uh um uh or might be Euro 42 000 Euros I think which is like uh a a Euro for every meter oh wow I really appreciate this because I see the way that poapp is changing from uh being just memories to also being um leveraging the power of people to do

good things so um we're experimenting with uh like when we do uh training guides or whatever like you don't have to have a poet but if you want a poet send a dollar that'll support the people doing it uh and so this is very much the same I'm I'm very happy to support left Terrace I'm very happy we're talking about it yeah and if nobody knows left Harris's uh proper OG and is very much at the center of all the uh Dao uh folk stuff and the Dow hacking is a major character in the cryptopians uh book and uh or a nice guy you you don't get much more OG I don't I like there is no more OG than left Harris so

I see my friends here that means that I'm I'm late so I better hop on yeah don't be scared by the German it's all fairly straightforward and the leftovers uh kind of explains the process um and you can donate in crypto so uh if you want a DM lifters uh you can um donate in crypto as well so awesome uh let me bust this stop share uh it has been great talking about poapp with you today we've managed to blather on for well over an hour so yeah yeah I uh same I'm got a lot to do today and I'm like do I

have time to do the Ben thing and but it's always so always a pleasure yeah all right let me stop the recording