ETH2 Staking Community Call #9 - ProtoLambda

Transcript

hey everyone welcome to the eatstaker uh community call this is the ninth call today our guest is proto lambda um you'll see him on screen with uh buddha so his name their combined name today is proto-buddha the eth staker community call is is an attempt for uh our community to spread information about ethereum two to make staking more accessible for the average user and to promote solo staking as much as as we're able to to do that we share information we produce these these community calls we have an

eath staker study master where you can learn more about ethereum too just a lot of ways you can get involved but yeah let me get into this today we have proto-lambda proto is a ethereum foundation developer if you've ever watched the ethereum 2 development closely you will see that proto is kind of the the glue or the fixer of of all massive problems um so uh proto tell us tell us a little about about yourself uh proto or d hook is my real name i just go by prototypes when she's here online

um and so i just through like kind of weird way i got into the foundation um um at first contributing open source um i solved many early optimization problems and i kind of just shifted through this more like this various researcher role where i'm not really researching as much as just talking to clients and hacking on various different projects within ethereum ii and right now i just think i've mostly just fit in with well more testing plus networks i maintain the attack net and then a lot of different client issues and network

issues i uh look into is it is it fair i i think of you as of the jack of all trades is that do you think of yourself that way or do you think of yourself differently yeah so i kind of just got like i i'm all over if too like i like so many different parts and it's hard to keep your attention in one part and there's always like we always need more people in devops or in these more technical things and so it kind of just turned out like well just do what is most effective at the time and just go try and back and forth and so far that has been really great i'm enjoying that a lot so i'm going to think of you less as a

jack of all trades and more of a firefighter you're the guy that puts out fires that uh that might be fair yeah sometimes i'm the one who puts up the fire or someone the one who becomes associates so like um right now i think towards main nets most clients they're ready in some respects there's just more testing refueling i also do a lot of reviews um and then it's just it's kind of networking stability uh where you can always use more people and uh i try to do my my partner can you tell us where you came from like how

how you ended up in computer science and how you it's a weird it's like a very weird story like some of you like i think what already knows some of this this i got into computer science first then well like i always liked programming i guess i went to this technical university in the netherlands studying computer science then at this exchange program during my third year of bachelor science in hong kong and i went to hackathons into different various things and i ended up dropping out just to go work at some country in hong kong to work

full-time on ethereum and well this was all planned i had so i thought oh okay when i get back to hong kong i'll start working there in the meantime at this open time or a good trying experiment i tried going to university for one more week i thought i have to try i can't just pause it forever uh but then again like i thought i want to contribute i want to do stuff with you i want to go to all the conferences there's no such thing as corona at the time yeah so i kind of like started contributing a lot and when i was noticed i got this big part into this bounty where you would program

the full specification at the time which wasn't really as effort into like a thousand lines of cards it's just like a code golf challenge to get it slimming down and look at the code like you'd have to know what's going on to try and actually do this kind of thing and then but all really well and then there was this conference in paris ecc very first the second edition and i got to meet some of the ephem2 researchers from there it got really started i wasn't even in hong kong yet but then i went to hong kong was lucky enough to meet other researchers there as well and then just just like only if to live

to every day talking discussing and writing tests and stuff online and then like just a few weeks into my first job i was hired for the foundation and stuff i was kind of messy to make that search but in the end i think it worked really well and i left some good smart contracts there as well so uh but now it's protocol all the time my smart contracts and why didn't you care about it well why didn't you care you don't want actually and just a few of them too so okay what got me started with ethereum one because just every briefly into bitcoin first didn't know anything about community or like

what the goals are or anything i thought well this is kind of makes sense to have money for yourselves very technical filming with cryptography okay i want to be in this and then i got more involved with meetups and so during the run up in 2017 it started with just a little bit about bitcoin then ethereum and more and more meetups and by the end of the year i was going to one or two meet ups every week and then go organizing some meetups and unfortunately later it crashed right was this dip meetups went down but i kept going i i'd like to hear him i learned a lot about solidity the epm the yellow pepper all that uh

and so i kind of just got stuck in here and never left i guess i'm enjoying it so far cool yeah um i love to hear the different routes that people get for the people follow to get into this space um it seems like and i guess this is obvious there's either the programmer route or the investor route um and they eventually find their way to being contributors uh cool right i think for programmers it's really interesting and helpful if they like for me i first did this open source work and then this bounty here bounties up right now like

anyone who likes to get into the space are really encouraged to just look at what's there and try and say hello online because hiring is first it's really difficult uh but it's it's helpful when you just show your yourselves yeah i found the same thing that at at first there are so many people talking that all of the voices are drowned out but if you're persistent and you continue working on the thing you're working on eventually you you become more recognized in the community um so yeah i i definitely follow that we have some really interesting questions here um before i get into those questions yeah

they're they're all secret now we have a published agenda um but i want to talk about uh i feel like this is an advertisement let me say be very clear uh east staker is friends with poep but we're not i wouldn't say we're sponsored by poep uh rather than that uh we're just sort of obsessed with the idea i'm obsessed with the idea of perhaps and i know a lot of other people like them too because every time i speak about it i sound like i'm doing a commercial that's actually my excitement it's not uh patricio giving me all of his uniswap tokens um so co-apps are an erc 721 token

they're a non-fungible token and they represent attendance or participation in some kind of event poep stands for proof of attendance protocol uh patricio is the guy behind it uh and i'm talking about it now because you can get a poet for attending this meeting pretty much everyone who's here has has been through this before but for the benefit of anyone who's watching this later if you watch our meetings live if you participate in our events live you can uh get a link that will allow you to go to a site poep.xyz or wherever it is and you'll be able to claim that uh nft token to keep it in your wallet now the neat thing is when you go to several events you end up with a collection of these

poeps that uh i don't want to say they give you credibility but they identify you as far as what your interests are and what you're participating in there's a very good chance that as as poep develops and continues on that reputation will be more meaningful and more useful to the community uh as far as like getting entrants into events um hey look at that topaz dev what's the con con so um yeah the poets are just a really

neat thing as an aside you can also if you choose to you can sell them i've seen them on openc.io i don't encourage you to sell them because i think that they're more meaningful as as a demonstration of who you are but there's nothing preventing you if you choose to sell them and i would encourage you to sell them for nothing less than a thousand each a piece so there's something there's something else about poep on on this event which um it's worth celebrating on on every event we have issued poeps and when i say we it's not only east takers but every other event that had pub and there were hundreds of them the pitch was always

these pop-ups are going to be useful for something someday they won't be like they are now which is like bitcoin something to look at and not being able to do anything else for the first time now if you get this pow up you get the ability to vote on the po up design of the bitcoin chain lounge and because the submission with the most boats is a wallet a thousand dollar bounty it may happen that someone is willing to bribe you for occurring boats so so there's an angle that really justifies getting this pow-up

unfortunately um gas prices right now are almost at the tearaway value uh we shouldn't measure them in gigaway anymore because they are almost over a thousand gigaway units so my advice is to not claim them right away you just input your ens or address on the field click claim and you're done because if you get to claim it on magnet it may end up costing like 20 or 40 dollars which is insane and on on a couple of weeks hopefully we should be minting pops on on the xdi network at least temporarily and and it should be free to do so so so

make sure you put your address on the on the field and you click claim and but don't bother claiming it actually so that's it for me and that's something that took me a while to catch um if you go to if you go to the the claim site you can enter and press claim just like patricio said you do not have to complete the transaction your your address will be stored in the database and after when xdi goes live which i believe it's very close now uh maybe in a few weeks because your address has been associated with that co-app in the database but you didn't claim it um it will be sent to you automatically which is something that

i guess even as an organizer i was sort of aware of but it it didn't really click with me till a few days ago and i don't want to speak for patricio but i know his heart when he says that someone might uh bribe you to influence you we're not we're not encouraging you to accept bribes we're saying that the potential exists and and poep isn't in anyone's control so what you do with it is completely up to you is that fair that's correct great and i i'm happy the east taker is so closely associated with poet because i believe it is going to i believe poep will be a huge part of our future uh

in terms of building reputation and a trustless system so i look forward to that potential um proto hey buddy uh so some some of the questions that i thought were really great what do you think about what do you think we're going to expect from adasha in the next couple weeks before main net launches anything exciting all right so it's this run up to mainnet now and like for mainnet we want to get the global things right and including some updates so it's not this balance between we test them on different test nuts or we're going to risk it in midasha

and test the full-blown life and so this is kind of exciting for the developers and for users i think it's really just like this final plan you'll have a large scale network to prepare uh get involved try running validators come to knuckle you can even try loads there like there's no like no no point in not trying like run it try it i think that's that's what i like now about russia meanwhile attack nuts um and compared to madosa i think it's like if you're more into it just for messing with the attack nuts and trying stuff you can get rewarded and you just go to

the attack nuts so check that out where are the attack that's hosted um hey looks there um right so attack nuts also there's this other challenge the dosha data damage which i also wanted to talk about so like those of these challenges you can find in the east taker discards or in the if to r d discards where you can find people talking about this working on these things and you can earn bounties for the tech yeah all right so i was talking about

this um if medusa you have the black nuts um in each taker discord if r d discard you could find a little hall and then for like less technical inclined uh folks the data challenge is still really interesting but you should know some coding maybe exodus as well we're looking for is these new insights taking explorer data api data and take a look at what can be done to improve the main to improve the test nets and polish this more towards maintenance one of the interesting things i've learned about uh the madosa data challenge we had uh laxman and steven on our last call is that a lot of the people who would do really

well may not really know a lot about ethereum yet and so if if you're watching this call and you know someone who's a data scientist or who's interested in data modeling uh it would be a great idea to connect them i know now that there's a uh there's a couple zip files that are circulating just with uh attestation data and just data right from uh the eth2 clients uh one of my good friends who who wasn't really familiar with eth is now picking through that and creating some really awesome stuff and the cross-pollination is that he's going to be eligible to win uh an award and he's investing in the

ethereum community so he's becoming more familiar with ethereum he's more likely to engage the community and we're benefiting from what he's providing so it's just an awesome opportunity cool um there's a lot of so security is is one of the big uh the big bears in the room and i know a lot of us a lot of us who have been a while don't been around a while don't worry too much about security like we secure it and forget it um and that's probably not a good a good theory but we secure it and monitor it but in terms of of ddos and using vpns do you have thoughts or

best practices for running client validators all right so there's this endless discussion about what is it but must hurt her and i can really like advise you with is try and manage your risk uh think about setup there are lots of great write-ups being made and i think like just as if the data challenge if you're motivated is this this is your thing uh you even want to help you provide these things to the community uh as a service brothers then just direct advice for myself is i think the minimal thing you should be doing as like the pure hobbyist stacker

is use some kind of vpn or have some way to be reachable from some other points as in like worst comes the worst uh you have some home issue or there's uh some target that those would be like absolute burns unlikely but like possible um this case is just rather easy to avoid and was probably like the cheapest solution you could get to keep sticking and have lost words uh you said you said i have access to a vpn i i know that there are you know there are two major implementations of vpn there is

uh v pinning vpn out to a service to obfuscate your location and then there's vpn into your home service as a as a way of securing against ssh exploits and definitely talking about the farmer like obscuring your your identity is not someone thought so does this thing are they separate the user layer all the identity stuff your validators from your network identity so if you are validating with your federator nobody knows it's you on this certain nodes you have to try to mine really hard it's also the data challenge uh privacy is this point of interest where everything that leaks privacy we should

we would look deeper into we're already really trying to push clients to hide as best as they can what's running and like avoid this kind of targeting there's like there's so many fellow directors that the problem is not being lost the problem is being is is targeted where at the right time just at the point where you think you need to propose a block that then then have a break slow and then you have this problem propagating a block on the network if you could avoid this with some vpn your ways ahead of most others okay nolan do you so i i was i'm always curious about your

perspective on these security measures have you thought about running your node through vpn um yeah i i was thinking about having one like run it on my ip and then if i need to i can flip it over to the vpn and if i start getting attacked there you can go to a different point through that vpn that was sort of my plan that's a great idea because i as proto's talking about i'm like i i really don't like some of the overhead that that comes along with running a vpn full time but running it as as a as a backup safety net yeah but and he's also talking about like like being targeted if if we're staking

from our our local ip it does so that argument is you know if someone knows your ip they've got you but ips are kind of transient you can pretty easily spoof your mac or reset your address um and people have been public i've been broadcasting ips since 2010 since 2009 when bitcoin launched like those are those are pretty much home or office ips don't you agree photo so it depends taking it's and especially if two it's about timing so like persisting this kind of attack is just ridiculous especially

like a small mobius who doesn't have that much to worry about it's not going to happen to everyone uh but then like timing and it just this short amount of periods where you know gets a little bit of extra stress that can be annoying so these kind of security measures you you have this network solution it makes more and more sense of our federation room it's not the primary thing to worry about but for security also just separating out your federator being like conscious about like how you like move around your keys and think about uh having a backup of your

uh of your notes and being able to uh move around the information of your validator this protection information so if you switch to different modes you're right ready from the get go to start validating again these kind of things they're also like i think for a home hobbyist someone who's running one or two validators is more critical to get right than the networking okay interesting michael i know you you have interest in uh network topology do you have you have any thoughts on this uh this is where i'm a little out of my depths in fact um so i did have our current topology right now

is we have like a load balancer in front of our validator so i had questions in regards to when you start running multiple instances so i don't think it's really an issue when we have a single validator but i was curious how you can because the logic being if i'm going to run a validator or multiple validators i want to as best possible mitigate any single point of failure so that being power and internet from the physical layer right so i want to we've got multiple lan connections from different isps that are coming into

a load balancer and that's just to mitigate the you know give us a form of redundancy but how does that in relation to the validators with upping the probability of being slashed because of how those load balancer works so i guess the ultimate question here is how do we solve for internet redundancies without the potential of being slashed or playing by the rules but potentially looking like like we're not sort of thing like that's the very minimal setup where you just have one failure to one week about that's like as minimal as you can go but then from there there's like ways to improve

on those threats both balancing being able to attach your validator to different backup nodes that's one way um then especially if these will remote beacon nodes then you're able to at least have through one big mod to perform your duties um so this this is thing if you run many photographers you could look into um then there are remote signer applications it's not just the validator you're running it so you take a remote signer where your keys are hosted so you have your keys isolated and then that application talks to federators and does talk to the commands that will be like the even more complete setup

something more closer to look like you know staking surface or like a bigger stake or whatever i feel like it's a lot more expensive to rent or borrow or buy or use your bot to ddos than the actual return you would get anyway so i mean assuming let's say i have 10 validators it seems like whoever is ddosing me is spending more to dos me than i would be making anyway so i i guess i know some dossiers can last days weeks or days or weeks but it seems like just based on economics that

the validators are in a superior position is that oh kind of so i wouldn't even ask things that i started off with is to think about risk management assisting on speaker if i go offline what happens and then what did you want to do is move your keys get online again somewhere like by no means there's something like there's no there shouldn't be any drama there it's just like you can be prepared for this um and exactly there's a certain cost to attacking someone and if that doesn't outweigh the cost of moving to some other different house there's no point in worrying too much

about it what you should be doing is think about your key management think about just general security i hope kind of does those vector bit more like being able to protect yourself from some kind of intruder or not have your home network light up right in the open yes your words make me think that one of the things that we could develop as a community is a organized list of priorities because i i find that there's people keep bringing up error correcting ram and i'm like error correcting ram is like the lowest worry of a priority of anything you could be thinking about

but for them at that moment because it's on their mind it is like the most important thing i it might be good for us to develop a rank of prioritization so that we can say well that's really below this and if you've taken care of the 37 things that are above it then definitely look at this but otherwise because i i do think dos is is likely to happen but i i'm also as you know kind of experienced in the sector i'm not really worried about it because it's it's not that high on the on the priority list for me right now right be more for worry if it happened to anyone like like everyone at the same time that's a serious critical issue

whereas if it's a small point of failure just a single note that has goes down but the rest of the network marches on the thoughts let's see it's manageable you can manage this risk and if at that costs are high that's taker should be able to move and protect themselves that kind of gets me to uh to hardware i'm sorry yeah did i cut you off i'm sorry your audio cut out um so i'll go on did you have something to yeah well can i add something just real quick again um

what do you think about doing multiple deposits in one transaction does it kind of reveal your identity as well so if someone does 100 deposits so yeah so you can take some precautions on this user layer where you make transactions or you put some gravity in your metadata blocks these kind of things or you're on f2 stats during the test nets there are lots of fun and nobody cares there may not you should be looking through this critically and thinking like okay is it more obvious that i'm running these validators am i exposing myself in some way and if you're not prepared to to be identified or have a feel bad

then you should try and minimize it as much as possible and have like a plan for what if a and b don't go unprepared into stacking so also um what about doing is it more is it smarter to do you one deposit per if one address or yeah it's just one example so now the whole sets are multis or privacy problematic things or if you're just making videos in the same if you're not address making them shortly after one another then it's kind of obvious that you're the same person and especially when you're when you're running on the same note it gets even

more obvious like all these correlations this thing that's just likely to be the same person the same machine uh it just makes the knife of some data analysis trying to look for this few attackers for these fuel validators easier and this making the lives of some potential attacker who needs is data also easier yeah talking about deposits doesn't the launch pad do all the deposits at once or is there some scheduling behind it now that's not the biggest concern but they should definitely do this better okay and then again it's like not the perfect solution either did you do them purely randomly and there's also just some

again if you can manage the people think of it as the same person maybe okay yeah try try and do these things uh without like exposing yourself like you do right now it's just it's just not the same i think definitely zero like with he just going into staking first and then adding these other functionalities later it's just good risk management have this if anything you can keep one like running stable and move into each tool

with this erp and that was just progressed we should talk about this as well so before that i was thinking it would be a neat uh neat feature request for tornado.cash to accept 32 eth as an increment so that people could easily maybe it would be like 32 and a half but you know because it goes from 10 to 100 and it would be neat for validators to be able to swap identities effectively with tornado.cache so this idea was around for a bunch of times in the during the eve denver hackathon i think it was named kettle core or something like that where they implemented the user interface for

tomato cash to do 32 year deposits it it kind of makes sense to at least think about your privacy and if you're in one and tornado is one of those obvious answers uh you were interested in talking about an eip right so did you see the erp there's a draft up oh yeah yeah yeah you know it'll be it'll be like lots of suggestions people nitpicking every little detail that's happening yeah that's uh it's it's hard because each of these little bitty incremental gr like madasha being released and uh planning uh spadina like

and you know the official eip for eth2 it all adds to the excitement and the potential that it's it's actually happening it it feels it really seems and i i'm not asking you or anyone to make an assumption but it feels like it's going to happen in november like i feel like we're we're on track for it so very exciting why are you laughing buddha it was one of the questions we expected or i expected at least let's be clear i'm not asking i don't i i don't i'm watching the same things everyone else is

what i'm suggesting is that all of these pieces are coming together um and it does a i didn't ask i can't thump you on the screen it didn't all right i'm awesome you could ask instead what are the next steps bro to after uh well first well finish the eip like there's a lot to do and emerge and it's not quite done yet uh good comments in the iep it's worth the discussion definitely um then from there it's a mix of coordination between clan teams they don't need to like say they're ready you need to think about it but i know it's but you know um it's worse than medusa for me i think

that's difficult then from there the deposit contract of course will try and enshrine the contract address will never be forgotten from there so it will go into the erp for sure then i think it's also just this this social effort definitely review all can help and everyone else to get this thing off the ground and ready for mine at um lunch then more feedback more test testing and then preparing it's gonna be like one exciting few months towards the lunch i i really see east daker

our role as promoting those individual stakers the people with 32 east who might otherwise have sent their deposit to a staking as a service host or to an exchange um and so as far as like getting this information out to people who aren't in this in the core of things but just making sure that everyone has a general idea how to get involved with the test nets and uh and be it maintenant launch so yeah great i can end up we have something all right so i think before the first test net then he was asking about how many clients we would need or how long run for i think it was

three clients for two months yeah my opinion here is likely the three and a half right so deco in my opinion is quite ready um pretty solid in refuel prismatic for so many users still like even if there they can be like ups and downs i think there will be more already uh lighthouse is very solid as well like i think just the three clients all we need is like more uniform distribution of users the better this one third split is the less risk than that costs this network is built around this two-thirds assumption where with two thirds it can have finality and it can be very healthy

as a network that's great so there's one client that can go down and we'll be on the edge right because with one more client things get even better like much much better if one client goes down you can manage if you have good participation and definitely count good participation at mainland compared to russia it's about real stake with like real high stakes just different from running a testing machine are you aware of any efforts to encourage so you know we had this this client uh medela madasha client poep challenge are you aware of any current efforts to uh encourage people to

use clients other than prism it's it's a problem everybody's aware of but are we are we doing anything or just encouraging so like the problem thing for to encourage people to run different clients has been great i'm looking forward to the results but it was already hinting at this time period roughly being i think with 300 000 slot mark was the official end then i can't wait to see these tokens being distributed and it's it's kind of exciting to have these this batch of testnet users and maybe based on their tokens heard their feedback or their motivation for genesis like this this will be an interesting time for sure like

how how these kind of does net users move towards maintenance and uh how do we go with um the voices will hear what kind of new setups we'll see uh news um explorer is improving and showing using more magnetic focused things and we'll be crowding the network to try and uh have a better look at client diversity yeah do you feel like i do that and i this you sort of alluded to this that the top three clients are rather interchangeable uh lighthouse nimbus and tecu uh that's not what i said

lighthouse prismatic and tekker that they are fairly uh at the same or personally i don't think they're at the same level of development but i think that they're more or less interchangeable is that fair i would say interchangeables and bad words i like them from others to deserve their own features definitely reasons when we choose one or the motor based on their situation but then again the other they should be all ready enough if you provide sufficient resources uh know how to run your clients and you're good to go um then the other clients they're pretty far i would say nemo's definitely is like a party they could stay three and a half or um i

think they can do it they can be in genesis but then again they went through a very rough uh road map to get nimbus off the ground and literally b2b this whole new networking stack there's lots of power level changes interferometry that are very challenging and they had a harder time with because they're they're doing things so differently right there's different programming language a whole new team um like uh whereas compared to prison not being running testness for so long it's uh definitely like it's a contender that's not the maybe not the very first thing you think of for

genesis yeah okay um maybe please so we had five clients how many clients do we want or expect on the main nets do we want to have all of them ready or just two or three i think like even two could work preferably three i definitely think we can do three at this stage or instead of shooting for like or you can do that i definitely support that first then again we should try you should risk manage your risks and think about security uh think about refusal the audits produced

so these kind of things there's uh it's not just about the first table also just like there's some sanity checks you should do and think reason through learning obviously of these clients that may be a good phil phil offered to have for example to have some diversity um [Music] two more questions looking into the future um now that with the we have one way of transitioning to ethereum two by depositing into the contract um let's assume that not everyone is going to be a validate on ethereum two

how is this the ether from one going to transition to e2 without being a validator is there going to be some kind of contract where you just expose it into and have the balance available on your right so economic activity from very get go from different risks and problems i'm definitely supportive of it i think it will be happening something uh during phase zero um and then again like it makes sense to start it and with just the minimal sense and scale up the features rather than try and throw more complexity at that problem but but is there a specific idea how to

move that ethereal different ideas a bridge is definitely still possible bridge especially is relatively easy to implement uh it is about like high stakes it can be implemented where like a very basic solution would be to have like a two week periods um if you're wrong with some kind of voting in the sides and if you're in two chain uh similar to how the firm two chain decides and leave your money chain um and then essentially you have a swell to lay bridge is one option it's just speculation um there's this demo of if you're uh 1.5

if um deco integrating uh gulifarium so that happens here is that in kotlin there is this short block processing and the actual contents of the block are processed on gully theorem not it's dedicated um and it's also just very interesting to actually see end to end and transaction go to ethereum too i don't think it's necessarily far off to just think about the complete picture and if that's if that roadmap is fast enough it's not if that's possible i'd much prefer that or for these workarounds like bridges yeah and i think many people also think

that since phase zero took so long that the other phases will take as long as well so what do you think about the uh the time between phase zero and one and one two one one pi and personally i'm foolish there i guess like the complexity are still of complexity and what phase zero solved is this networking and database parts of the clients which are very very complicated to move the complete networking stack the concept of stations there's uh this solution is designed to aggregate attestations uh propagate aggregates this on a network level

it's a huge step and it's just that users don't really see as much of this and then for example they're these committee management subnets they're already there there's just no no data to photo movie subnets but they do exist and this structure is definitely in place for facebook uh to make quicker progress from there um so yes i'm positive there that we can improve and not take longer than phase zero at least yeah yeah i'm afraid it would be confidence killing if it took as long as phase zero but it is it's kind of nice to see l2

is stepping up so maybe uh maybe the pressure on l2 to step up is is okay and and things are going to to balance out i don't think that we would have had as many layer two solutions if beacon chain had been launched before now because everyone would see the potential of sharding and lay off of l2 one thing that i wanted to catch before we finished was hardware um i i think you run well if you want to tell us what you run but a lot of people are curious and we talked a little bit about the network stack

but in terms of client stack what what do you think are minimum average and best in all right so it's complicated so um generally i live pretty lightweight so i don't have expensive home hard turn and what i do manage is like lots of different modes on the russia for testing for the deafness or the attack nuts where it's if the researcher is more complicated to just keep up with all the devops um but i think my experience so far but for the business skill that's not as large as with russia

it's i think there are now 55 other directors in the registry i think it's advisable like it's a good idea with like a 4 gigabyte ram kind of setup with at least two if not four currents it seems like a reasonable pick for me like conveniently running legal notes and my notes are actually running on lower specs and the current attack net and on the current russian network now for some clients it works well enough christmas lighthouse can do it although you're really just pushing the limits there so just like have some room outside for gigabytes where i mean

more like i think about it for prism for deku especially it's a good decision to go beyond that um but for gigabyte assumes that we have no finality issues right issues you have this problem where there's different choices where the client needs to deal with more data it can't be persistent to the disk disc memory just be able to switch more quickly these kind of things there they hope memory and you need times which are with larger test nuts and low finality bigger knobs definitely helps

and those are running and each one knows requires way more than four like if eight gigabytes is like limits yes especially just for and stability for training to optimize costs which see more optimizations actually still working on you have been optimizing state a lot i've been working on ourselves more so saying speed and database i think we'll get there and we get more of really with our okay yeah the last question i don't know

if you have something else but no i i was actually looking to open it up to anybody else who wanted to uh let's imagine the worst case scenario we keep staking for two years everyone is happy and then we try our withdrawal keys and they don't work but is there any solution that's i don't know i think definitely before genesis this is withdrawal methods channel r d and if you're discord from a topic discussion i think it's it's good to talk about this kind of thing through the community and kind of this problem here is that you want to design it really really well

it's kind of withdrawal herder joel method it should work it's really simple it's just not like integrated to the theorem as well just not to withdraw interfere which like right now we're just awaiting facebook manfest to make that really effective i think maybe we'll deploy some kind of test suit for this uh i think this is certainly worth it and then again right now not the priority for before lunch definitely definitely if we take the route we have right now and let's say we launch and everything is fine would we need to fork to do some some forking on the e2 chain in order to make

the funds accessible or is it yeah so i expect that if we're going to fork leave to chain regardless and we'll definitely include things like this where we process withdrawals or recent effects all right i think the biggest concern yeah and okay so if say we launch e2 and it all breaks or say we have like a big finalization issue like we saw on modasha would that be something we just power through or would we relaunch using the same deposit contract i think

it was unfortunate with modashia that we powered through and clients harder than what that improved there it's hard to imagine like things going like that again and even if they do like put stronger out of this or power through it more quickly and then worst case of the worst it could be a race turns then yeah like with little economic activity from the sterns good coordination and expectations setting expressions from stars i think like worst case is still it's nothing like slow defense either so one of my favorite new questions i've

just been thinking about for a while what is the thing that you're interested in talking about or that you think about that we haven't really breached today what's whether even if it's naughty too related like what is what are you obsessed with right now well all right um i think like right now i'm staying here via uh you can thanks for our team um i'm getting more and more interested in all this kind of data about business and probably finding more people on difference and people are kind of interested in helping find these issues in test nuts so open tools to anyone that's

kind of like once you get into this uh we're out here we'd love to talk to you and we don't make that change it's like one way to get started but then if you are interested more networking or anything else there may be some spots or something uh interesting for you yeah i think you're not talking about kind of like growing this ecosystem of tooling and information do you support all these new features or the big 10 exposure um it's kind of important just before lunch to get this house awesome so yeah i guess that is the

closing like we talked about we run these calls to onboard people to give them more exposure to the ethereum ecosystem and especially e2 um and so that the madasha data challenge is a tremendous opportunity um coming into the forums to learn more about how things work one of the the best ways that i found is just creating a little bit of your own content like looking for something interesting that you don't understand and exploring it for a while and then producing a text post or a video or a podcast about hey this is the thing that i didn't get and i'd like to tell other people what i figured out um

so yeah i really appreciate you coming today to help us all uh hear your perspective i'm i'm always so impressed with uh the quality of your work and your solutions uh so you're one of the people i admire in the space so uh it's great to be able to talk to you yeah yeah i really appreciate like i'm not as active in the chat than with staker i've definitely learned a lot and where's these calls so i've been great i appreciate that thank you great uh and thanks everyone for coming don't forget to claim your poep um actually don't forget to go and enter your eth1 address so when x die scaling goes live in a few

weeks you can get your poep delivered for free and i think it's time to go check on the price of uniswap all right thanks everybody for coming up thanks for watching bye