Shanghai/Capella Community Call #1

Transcript

foreign [Music] [Applause]

thank you

Tim Beiko: um good morning afternoon everyone

depending on where you are uh First chappella Community call um so I guess uh these are are helpful to just like answer people's questions about the upgrade um we started doing them with the merge and I given the introduction of withdrawals um yeah it makes sense to uh do them as well so that validators and and just node operators and infrastructure providers could come and ask questions um didn't have too much on the agenda today I think uh if if we have Danny and myself here we can maybe both create a quick overview of like the entire upgrade so you get withdrawals but there's a couple more things as well

then maybe go quickly over like the withdrawals FAQ um and and just like show that it's there and what type of questions it answers and um if folks here have questions we can spend most of the most of the call on that um but yeah I guess that maybe to kick it off uh so Shanghai obviously as I just said the biggest biggest thing at discovering are withdrawals and we can we can dive into those in quite some detail uh after um there's also three other changes that will be included so uh coinbase address uh will be repriced when you access it um so this is just a small gas gas cost change um push zero is a new up code that's

coming in so uh which as it as the title says pushes zero onto the stack um so you can get some gas cost savings by using that if you develop smart contracts and then um maybe the change that affects a smart contract developers that the most is um this limiting and uh gas metering of init code um so previously or I guess currently still on my net uh edit code is not uh limited Beyond I think the entire contract size um so this adds a specific cap for debit code and also adds uh gas costs uh for every uh basically two gas for every 32 byte uh of 32 bytes of init code um in a contract so that will change

kind of gas pricing uh with regards to um but yeah outside of withdrawals that's what there is on the El I don't know Danny do you want to quickly share on the CL yeah

Danny Ryan: so on I will do the non-withdrawals um in the consensus layer there is what's called a double batch Merkel accumulator essentially you take uh blocks block routes and state routes and you put them into lists and then uh every once in a while once those lists get full you hash you do a Merkle root of each of those lists and then miracleize those and so you have a

single root and you put that into a list itself the these initial lists can be cleared um so they don't grow unbounded uh whereas this like double batched list does grow unbounded but like on the order of 10 kilobytes a year but this allows for like succinct um historic access forever this exists it's been uh in the beacon State since phase zero but because of the way the state roots and the Merkle roots are kind of like hashed in the single list um if you want to prove something about blocks you have to have like a little State Route you have to have this additional route to prove things about

blocks and if you want to prove things about states you have to have this blocks about State um so because of that some of the historic serving that we want to do against States is not quite as clean as it needs to be thus um that's being refactored and revamped a little bit such that we have two um we kind of like keep those two pieces of information the historic summary that that batch of block roots and that batch of uh State routes as independent instead of Smashing together into a single group this was led by yasik and should help with succinct serving of historic stuff cleaner serving of historic stuff very technical it helps

with you know State sync and other things like that that probably as an end user you won't end up um really really having much fun I think that's the only change other than the withdrawals which are obviously uh one of the main reasons we're here

Tim Beiko: yeah um so do you want I guess Alex is not here but yeah do you want to give maybe an overview of I'm not sure if it's like the right the right

Danny Ryan: yeah I talked about like the validator process and how that we're kind of that's being extended and then get into the technicals from there people have

questions

Tim Beiko: yeah let's do that

Danny Ryan: so as you know validators can deposit become activated there's an activation cue normally it's pretty empty sometimes it can be long similarly today already you can exit and then if you if you initiate and exit you go through um the exit queue you're not done validating you still get roles and duties until you're on the other side of that execute normally it might take on the order of like four epochs it could back up on the order of months just like half the validator so I was trying to leave it once um so you can exit today and then your your validator is no longer active uh but your phone's kind of sitting limbo so what this upgrade primarily does is completes the circle and brings allows

for uh both kind of the principal there to each balance and the validators rewards to go back into the execution layer back into kind of normal user layer normal ethereum accounts out of the beacon chain and there's two primary methods you know I talked about how you can exit so obviously once you exit you probably want your uh your eth back so all of your balance uh 32 Plus in your Rewards or 32 each minus any penalties once you get out of the execue and once you go through the men validator withdrawability delay which is 256 epochs which is 27 hours so exit Q Plus 27 hours now you are withdrawable

and on some sweep uh there's a certain number of validators 16 per block that can go back into the execution layer it's around Robin sweep your funds will eventually be picked up and put into the execution layer there is another method of withdrawal called partial withdrawal and this is if you are active and have balances in excess of 32 each so you have some rewards on that account the rewards in excess of 32e in that same round robin process will be sweeped and swept and put into the execution layer that's the mechanism at its core there's one big caveat and that is there are two

types of withdrawal credentials if you set up a validator you know you have an active key and they withdrawal credential at Genesis the only available withdrawal credential was called the Xerox zero zero BLS withdrawal credential so you have a BLS key essentially that controls the ownership of the funds not the active not the active key if you have such a key you don't have a essentially you haven't Not Elected a destination in the execution layer for your withdrawals to go to so if you have a 0x00 key initially you will not be a part of partial withdrawals and if you exited and you were withdrawable fully out of that queue and passed that 27 hour window your funds would just kind of sit

there so there's a new message type called BLS credential change vlsg BLS change operation something like that vlsb change operation I apologize for not having that on the top tip of my tongue but there's a new operation uh that tools like the staking CLI and eth do allow you to essentially convert your BLS withdrawal credential into an execution layer withdrawal credential so it goes from 0x00 version to zero zero one and essentially you say one-time operation in which you sign a message with that BLS key to say what I want the ownership of these funds to be on the execution layer

so it's essentially taking a BLS key and electing permanently your execution layer address your normal ethereum address that funds will go into um so once you do that if you're active you will now be a part of that partial withdrawal Suite so any balances in excess of 32 each will go in there so you know if you have a leaf that you've earned over the past couple of years um you don't have to exit to get those funds you just have to do the change operation if you don't already have that type of key and again at that point if you exit and become withdrawable your all of your funds will go into that execution layer address similarly if you're already exited or if you do the operation in the other order you exit and you wait for a

while and then you do your BLS change at that point you'll be part of this week so that's that's probably one of the very important things to understand here is that your validator will essentially continue to operate as it is today if you have a BLS credential until you um elect to sign that message and so a very important thing is kind of the ux and safety around signing you know you can use that staking CLI tool although you have to use an upgraded version that you may be used to deposit there's also a great tool called if do that a number of security Auditors are taking a look at right now as well that we've been using and testing and as a nice ux to do um and then the the step from there is

to go to your client of choice that you run you know your node software and there will be some sort of interface some sort of uh tutorial some sort of guide on how to take that message load it into that machine so that it gets broadcast to the network and uh picked up into a beacon block

Tim Beiko: question there's a question from Trent in the chat are there any good tutorials um that people can follow to do this

Danny Ryan: great question I was looking at the new staking FAQ which we can take a look at

in a minute um which does link to I believe it links to how to do it in these do and then it is cataloging the docs from the various clients on how to then take that message and put it into a client I believe we might be at like three or five or four or five clients that do have this documented and that's a very high priority for them to get over like the next week the staking CLI last I checked yesterday there is a work in progress PR to add this functionality and I would suspect that's also done probably next week so the staking CLI uh tutorial and guides will also be updated as that's completed

Tim Beiko: okay uh two more questions so one from

YouTube um and these by the way don't all have to be for Daddy if you're on the call and you have the answer please feel free to jump in um but uh yeah from YouTube I recently came across eip4736 consensus layer withdrawal protection can somebody please explain this a little bit more

Danny Ryan: I only have a cursory understanding of this I would um you know if somebody else does please jump in because I cannot do it justice at least give the background um there's certainly some fear that uh some people might have their keys

compromised uh whether they you know exposed them in the clear somewhere or some sort of mishandling in in key operation um in the past couple of years and so there's the worry because you have to do this change operation there's a worry that you know there's a race between you and an adversary and somebody who hacked your keys and if the adversary gets their message on first then you have um you know they essentially get to finalize the stealing of your funds that maybe they've been trying to for a while um I believe that this EIP is a standard to do

some sort of you know identity verification with large operators to to essentially put your key into their pool early rather than hoping you know and racing Against Time that said I do not you know this is an attempted kind of social solution to try to give the correct owners of funds rather than the hacker the ability to change their their key um but I do not know how the details of how they intend for this to work

Tim Beiko: yeah and Puja uh shared in the chats there's a peep and Eep uh about the CIP that's happening uh in two weeks uh and it should be live uh a week after that's

what they're recorded um and so for the people who don't know ppdeeps are like hour plus long deep dives into specific eibs with the authors so um yeah keep an eye out on the ethereum catharters channel uh for that when it comes out um Christine you had a question about public dashboards tracking the withdrawal credential updates for the withdrawal credential sweeps um there was uh answering the chat Barnabas do you want to maybe give some some details there

Barnabas Busa: sure uh so the current uh stage is that we're working together with uh we contain and block Scout and they both

are going to be able to provide a nice visual interface where we can see which validator has withdrawn or which has not and also I mean we can change there will be a way to see how long it's going to take till the next sweep this could come into your validator so it is upcoming and we do have a working Beacon chain right now for the running devnet so it's the latest is running on it that's where all the development is happening right now

Tim Beiko: awesome um a couple more questions from YouTube

uh so I'll get rid of an order is it accurate to say that the withdrawal message expires after two hard Forks

Danny Ryan: this is that um changed the spec the other day that Marius didn't understand it's very complex I've just spoke in front of you Mars the answer is no this was a a change the spec recently um to essentially sign the message in a very generic way uh with the Genesis work version so that is valid for all Forks um this is a PO as opposed to how like say attestations and even exited exits are signed where they're very for specific and they kind of degrade across

Fork boundaries so the good thing here is once you once you have a tool it will work period

Ben Edgington: uh Ben so Danny that's covers the credential change message I think we agreed to use the Genesis full version but the withdrawal message remains signed with the full version right so that expires the exit

Danny Ryan: yeah and I I know the terminology is kind of no but but you're right the message I was talking about wasn't a withdrawal message as well so there are exits and there are key changes if you're exited and you have you on your keychains you will be withdrawn so so yes then the the exit which is actually maybe a more akin

to a withdrawal message uh does degrade over those two epochs and I don't suspect that something that is changed although it's been discussed it won't be for this work um it's something that's kind of been discussed latently here and there

Tim Beiko: okay um okay another question about the the two cues and this is something that's like comes up I get a DM about this probably every day um so can you please cover in what order to withdrawal q and sweeps are processed so I think basically the I think the interaction of the exit queue and the withdrawal sweep we

probably need like some sort of animation for this but that idea yeah right yeah

Danny Ryan: there's an activation cue you are active then you submit a voluntary exit and then you enter into the exit queue the exit queue is you're still active while you're in the execute you still have duties you can still make you can still uh gain and lose rewards based off of your applications and proposer performance on the other side of the exit queue you are no longer active but you are not yet withdrawable there is the men validator withdrawability delay it's 256

epochs that's about 27 hours so you're out of the exit queue you are exited you then have a 27 hour delay and now you are withdrawable at that point you're part of the you can be part of the withdrawal sweep which is round robin so you know if the last if you're validator 32 and uh on the last block it left off at 31 then you're the next one in the suite if you're validator 32 and the uh last block it left off was at 64. you're almost at the end of this week it's gonna go all the way to the end and then back to the beginning so exit queue withdrawal ability delay then round robin sweep

clear as day right yeah

Tim Beiko: someone says we should call it limbo when you're exited and waiting to be to be withdrawn

yeah

Ben Edgington: yeah it's not it's not very long limbo right I mean even if every uh even if every validate has zero zero one credentials which they don't at the moment only about uh forty percent do um it will still take I think four days to complete a whole sweep through the entire validator set so once you are exited you've got a maximum of four days and probably much less uh to wait until your funds appear right a maximum of five days because of the men validated withdrawability delay which is 27 that only applies if you've only just

deposited right um once you've been running for more than 27 hours that that no longer applies if I remember correctly

you might be great I need to double check

Ben Edgington: yeah it's just some people um becoming active and then becoming inactive straight away um yeah but once you've been running 27 hours you're good

Danny Ryan: I think you're right

I understand laughs

Ben Edgington: I I wrote about it somewhere you can you can you can take a look

Tim Beiko: um one for the El side as then oh you think you're wrong that was quick

um the withdrawable epoch is set as the

validator.exit EPO plus Min validator withdrawability DeLay So I believe that uh you there is the the delay there I said corrected

Tim Beiko: okay next one uh Delight clients because you dare to answer in the chat so we're gonna get a full answer out of you and the question is um is uh can you provide information uh that explains the new block format and how to detect withdrawals on the El um so uh yeah I like kind you maybe want to just give a quick overview of how the Els block how the El blocks change with withdrawals and maybe also like

um this idea that like withdrawals are process there's like an operation and not a transaction itself

Lightclient: sure yeah so you know the yell block header changes to include a commitment to all of the withdrawals that go into the block and that commitment is very similar to some of the other things we're committing to like the transactions in the block and the seats on the Block so that's just adds the block header and then the actual block body that is sent over the peer-to-peer network is extended to include the withdrawal objects this just lets us get them from our peers when we're sinking so that we can store them and

and uh send them over the Json RPC if someone requests information about them but they generally originate from the CLS over the engine API they are just added to the the calls that see the El with a new new block information so it just sticks in to also support that withdrawal format the withdrawals are slightly different than transactions because they are operations and they are uh not really validated much of the El layer they're mainly just given to us by the CL we verify that you know the withdrawals root matches the same thing

in the execution payload but we don't check any of the conditions about if the validator is properly withdrawn or at the index of the withdrawal is correct and so we just process those withdrawals exactly as the seal gives them to us after we process all the transactions and and the block

Tim Beiko: um thanks we'll see if there's follow-up questions about that um next question was there was like a couple of questions on like test Nets and like the order and timelines um I think like the higher level answer is like we haven't made a decision yet um so I'll take all of this with a grain

of salt but generally um what what client teams were thinking about is um we've launched a bunch of uh of devnets and and are still launching some uh to testings I think when things are a bit more stable probably in the next week or so uh we plan to launch a longer lived new test net to allow people to uh to test withdrawals and other Shanghai kapala stuff on it um kind of like we did for the merge we had a kinsugi and Kiln um and then I think a couple weeks after that assuming we don't find any issues that things work as as we expect them to uh we would move to forking the the different uh existing test notes so Gordy and sepolia

um we don't quite have a decision around like which one we would Fork first um sepolia is generally easier to Fork because the validator set is is closed um so anyone can sort of uh sorry not to anyone can be a validator on sepolia and it's easier to coordinate that fork um that said Gordy is like a better test for mainnet uh because uh homesteakers can also uh stick on guardian and see how things go there um so which of those two comes first uh I think we'll we'll need to decide but we'll have them happen usually within a couple weeks of each other and then assuming those go uh go smoothly there's no issues then I will schedule my net uh for a couple weeks after that

so I don't know how that lines up in terms of of dates but I think in terms of just like the steps we need to take um yeah that's that's basically it launching a new kind of Long quote-unquote Live devnet uh that people can try 14 sepolia and Gordy TBD on the order and then once all of that is stable then we would move to 4K magnet um another YouTube question was is there a way to access or retrieve the complete withdrawal Q State including your position in the queue for a full withdrawal on the ethereum blockchain

Barnabas Busa: Explorer will keep track of it so you

will see when uh it's going to be your turn okay

Tim Beiko: beyond the Explorer yeah the node doesn't track it by default

Barnabas Busa: yeah but the BLS pool is public so DBS of every node is gossiped so as soon as you publish your own uh BS changes and it's going to be the same kind of way uh swept in the same fashion as the partial

withdrawal

Tim Beiko: sweet uh next one was uh someone said there's a translation subtitle function in Zoom that we haven't activated I did

not know that this was a thing but yes I can look into it for the next call uh for sure um if you want to send me like a link about that I can try to see if we can we can activate it okay uh no other questions in this chat um there was one more on YouTube uh can you please elaborate a bit more on the change from 0x0 to 0x01 since 40 already have 0x01 set would it take uh would it take four days to change everybody else uh to the credentials to cycle through the credential updates

Danny Ryan: yeah I believe there are 16 of these operations that can be

included on each each beacon block so each non energy slot which generally you know that's 99. percent of slots uh so yes it'd be 60 of 4.3 days uh would be how long it's going to take so a few days and that's you know we don't know how many people are going to be doing this immediately before um or doing it over the course of you know a few weeks a few months so so you know we we shall see

Tim Beiko: different uh withdrawal question uh so um basically they're still interested in allowing exits using only withdrawals

credentials uh there was discussions about this on forums and GitHub over the past year and this is something that would be uh helpful or critical for uh liquid staking providers um where you can initiate exits from the anyone have an update on that

Danny Ryan: um I know there are a couple proposals in progress to do 0x01 um you know withdrawal credential triggered exits um and so we could have Smart contract triggered exits which would um help make a number of these like different pooled operations more

trustless my gut is that there's an appetite for these but there's not a full proposal nor um you know a governance kind of discussion around that yet

Tim Beiko: got it okay and I think that's pretty much all the questions at least I see right now so maybe before we have sort of client chilling um it's sort of going over this specifically we have this new withdrawal FAQ um that uh the devops teams and some other folks at the EF put together um I won't read over everything but it

does cover like pretty much everything we've discussed so far as well as much more technical details um so the URLs no side ethereum.org at Launchpad withdrawals FAQ it's linked in the agenda for this um I think this is probably the main thing like for people who want to understand the process in more detail without reading the actual spec line for line um I think this this gives it a good overview we'll keep updating this in the coming weeks I think uh if there are questions that people would like to see here but that aren't there um you can either post them in already

Discord um I mean if you tweet them at me I'll make sure that they get added here but the goal is that over the couple next couple weeks and months um we uh we had this as like the comprehensive FAQ for withdrawals um let me check the YouTube actually I can't check the YouTube chat anymore I'm sharing this but any questions about the FAQ itself oh okay oh uh there's one more question on YouTube is it accurate to state that

slash validators leave the queue before those who request voluntary exit the exit queue

Danny Ryan: no if you're slashed you're put into the execute but you can no longer receive rewards um even if you continue to try to do your job and will lose money while being in the exeq uh so no the exit you cannot be subverted by slashing and then the once you become withdrawable which if you're slashed I believe there's a much larger delay put on you because there's kind of like this trying to see all the amount of people

that were slashed in the same time period but once you are withdrawable um any remaining funds are put into the normal round robin sweep so it will take guaranteed longer to get your money out if you slash yourself than if you do a normal exit yes

Tim Beiko: Barnabas do you want do you want to answer do you want to repeat the question

Barnabas Busa: yeah yeah the execute is six months long oh you're gone yeah if this exit queue is six months long then you get attached then you will leak it till the exit queue clears so that is not a good time to get slashed

Danny Ryan: correct and that's one of the primary reasons it is that way because if the execute was really long and maybe people hadn't been slashed recently you could try to like subvert the system and kind of get around the queue through that mechanism so it's certainly supposed to be not incentive compatible to take that path

Tim Beiko: uh sorry reading the chat here okay anyone have an update on Mev boost

basically and yeah I guess none of them are on mirrors yeah please

Marius van der Wijden: I think from what I've heard Mev boost will be able to uh like we'll support the fork and they are working on uh like supporting withdrawals and and will hopefully be ready by the uh by by default page so yeah they are they are working on it but unfortunately they are not part of this call and also yes there's a disconnect between

the core devs and the flashbots team yeah yeah Enrique which is like the maintain like one of the maintainers of Meb boost yeah

Tim Beiko: we can try to get some Mev boost maintainers at the next one of these and and go over it in more detail for sure yep

Christine Kim: do you guys think that um like is there any like youth core devs focusing on Mev or is it really just kind of like the focus of flashbots I'm wondering if there's kind of like people dedicated to Mev on the e-core dev side um aside from like research stuff like PBS but more like actual implementation stuff like is that a priority in your guys's mind

Tim Beiko: oh there definitely are people so I mean you know first I think the one of the biggest stuff like Thomas like the CEO and founder of nethermind is also part of flashbots um and has been for like uh several years now so I think um there's and there's other developers you know um I think like clients I believe might still be a maintainer of Mev boost if he uh if not he was definitely

one in the past Alex is

Danny Ryan: Justin tragg is yeah there was a number of there's a number of maintainers that are not flashbots on map boosts um you know rig the robust incentives group spends quite a bit of time not just thinking about PBS but thinking about issues that emerge from Mev

um I think in terms of like the development right now primarily Mev is handled through that Mev boost mechanism and so I think uh there are touch points from the core developers on that but because any sort of like native L1 mitigations or not what's happening in the next couple of forks as far as I can see um you know there's not a ton of depth focused on it right now

another mind

Marek Moraczyński: for example we not only have client developers but also Mev boost maintainers like Sarah Liu for example so we are working close with

flashbot steam

Tim Beiko: yeah and yeah uh definitely we can cover

this in more details on on the next haul any other questions

Christine Kim: this also isn't a withdrawals question but I'd be curious to know kind of um updates around like client development um it sounded to me like the eof stuff was um sort of moving on the eof stuff but like the uh like the harmonization I think between elcl block headers was like partially to support like clients and wondering if like there's other blockers for like client development like what's

the status on that can any of the youth core devs like update on that on like a summary of like like client development and um some of the the protocol level changes that need to happen to to me to make to like I don't know help that

and I should have uh miris sure so uh gath

Marius van der Wijden: used to have a working like client for a really long time uh which was not enabled by default but which was kind of supported the problem was that it it wasn't compatible with the merge and uh so there has been a big initiative to make The Gaslight client

compatible with the merge um one issue with the with the gut Lite client is kind of that you you have to trust the light server so like you have like there are people that run light servers and the light clients connect to those and you like it's not a trustless relationship uh between light server and like client um this is something that we can only really get with uh with worker because otherwise it's it's kind of too much computation to actually verify everything but yeah with the with the new so with uh we

have Beacon chain light clients and we have now also a gas-like client and the idea is that uh you can run the beacon chain that client with The Gaslight client and uh everything should work fine we have a working demo but it will take uh I think some more weeks maybe months to to get it in a state where it's actually uh usable well but the community

Danny Ryan: right so Nimbus and loads are I think are very active in um specifying and developing on Big Chain like clients

portal network uh is in a very kind of like Alpha State as at the end of last year for serving um historic execution layer blocks but doing Integrations that help with um trustlessly getting the data needed for Beacon chain like clients is a high priority for them this year additionally um good smart contract Beacon chain like clients are very high value item you could kind of use the sync committee construction pretty easily although there are a couple of teams that want to do a much higher security construction and are attempting to ZK either the entire Beacon chain or at least the entire FFG core kind of

consensus part of the signatures so I'm actually expecting some pretty like good like client smart contracts this year as well in terms of the harmonization effort that's an important effort to kind of reduce this uh the case in which I've lightly synced to the beacon chain I needed to insert a block header into my execution layer it would be much cheaper and kind of easier to do that with the harmonization although it's not impossible to do so um and it's still in like a trustless manner today you just kind of have to download more data so that's an important I think iteration but it does

not block like clients from coming online

Tim Beiko: um any other questions if not um we have a couple client teams here um do you all want to take a couple minutes to talk about what you've been working on for Shanghai capella why people should use your client anything special they should be aware if they're already using their client yeah

right

Marek Moraczyński: yeah I I can start so you can try another mind if you want to have a very quick synchronization or we've made

it only a few hours uh if you want to support uh execution client diversity because we are still minority client uh you can use half check plugin which is making staking even easier and help you with monitoring the node you can write your custom plugin uh we will we have a good performance if JS with Json RPC and generally I will send you a couple of links

Tim Beiko: cool any other team

always happen to Shield taku please do

Ben Edgington: uh good we're in good shape Shanghai uh already I believe that we are doing most of the heavy lifting on the latest devnet correct me if I'm wrong but I think Tech is running 75 of the validators um and uh looking strong um a couple of recent things um we finally got doppelganger detection implemented so that will be coming out in a release very soon um bring us in line with the other consensus clients and coincidentally with Nimbus who works uh on exactly the same feature completely independently um we are forgetting historical consensus blocks so we have a new minimal mode which means that

um once blocks get uh outside the minimum mandatory retention period Then you you no longer store them in the database so that keeps your consensus client database down to I don't know be something like 30 30 or 40 gigabytes um so very low footprint uh on that is optional you can store the history back to Genesis if you wish to do so um yep so that that's

uh let's take you

Tim Beiko: any other team here okay Marius do you want a few moments for

death

I

Marius van der Wijden: I don't know I think it's it's very important to have plant diversity um uh but it's also gas is a pretty good client and uh we are in a really good spot with regard to withdrawals um yeah so run our client if you want a really good client but also consider running maybe some of the Lesser known

plans

Tim Beiko: um okay I think that's pretty much everyone and oh actually one final thing before

we uh close out Puja do you want to talk about the the node operator survey

Pooja Ranjan: yes thank you Tim for that so as we know that client diversity is one of the important focus of ethereum blockchain we would like to have a diverse Network share on the blockchain um even Marius just mentioned get is doing great but we would like to have other clients considered for running ethereum loan so if you are running ethereum node whether it is on execution level or on the consensus level please consider responding to this client diversity survey this is a way of

collecting your thoughts and we will share this with the ethereum client teams so if you have anything to say positive about the client do mention it if you think that there is a Blocker in the present plan that you would like to be fixed so you can adopt that and run your ethereum node please don't mention so yeah it would be great to hear from you thank you for taking our time to consider this survey

Tim Beiko: um anything else before we wrap up okay well yeah thanks everybody for coming on and um yeah we'll probably have another one of these as we get we get closer to uh upgrading test Nets and and whatnot

yeah see you all soon

bye-bye thank you thank you

Speaker names adapted from ethereum/pm, CC BY-SA 3.0.