Community Call #15 w/ Danny Ryan

Transcript

[Music] [Music] [Music] so

[Music] [Music] [Music] [Music] [Music]

[Music] [Music] [Music]

[Music] [Music] [Music] [Music]

[Music] [Music] [Music] [Music] [Music]

[Music] [Music] [Music] [Music]

[Music] [Music] [Music] so [Music] [Music]

[Music] [Music] so

hey hey everyone welcome to uh the east staker community call number 15. uh we have our who must be our favorite guest danny ryan with us today um before we get into it i'd like to introduce everybody uh we actually have a lot of a lot of our notable friends here today but with me today is patricia worthalter who is off camera right now but he represents um the poep program the proof of attendance protocol um and everyone who attends today either on zoom or on youtube will be able to collect the poep but we're not going to make it too easy

for you and we'll talk about more about that later uh we have buddha you want to say hello buddha yeah sure hi guys uh nolan is uh lamboshi as you call him are we gonna are you gonna be nolan or lamboshi this year we'll see whatever depends on the price yes and uh michael who you know is in vetica hello and our guest today is danny ryan you've done this with us once or twice before i was on a call where we got like raided by uh and it got ended very quickly and i

don't know if i was asked back so that was one of our very first ones wasn't it yeah i think so yeah that was the learning call because that was when zoom bombing was all the rage because everyone was at home doing nothing yeah that was um nolan and i refer to that as the call we don't discuss [Laughter] but yeah so i'm i'm delighted to have you um danny is uh an ethereum foundation researcher who focuses on uh ethereum two and uh what we recognize him is is kind of the the herder of the implementers he's where the view of uh research and implementation

yeah i lag out it's what i do it's kind of my hobby um danny has the best view of research and implementation and so he he might be one of the people who is most able to answer our questions about how things are unfolding so welcome to everybody we have a bunch of questions and uh we're going to get rocking and rolling so i think the most important thing that we want to get out um the beacon chain launched december 1st which is almost six weeks ago um what can you tell us about how that process went and how it's going right now so went well uh first and foremost i i think it was a resounding success and

that was uh both due to the astounding work that uh researchers engineers have put into this for years and years and also the community you know the hobbyists institutional players like everyone stepped up everyone got those nodes running and we've had like a pretty fantastic stability especially in comparison to some of the test nets that we saw it really does matter if something's on the line it seems like people really want to keep their nodes online keep them running well and it's been a relatively straightforward task given that we have relatively robust and stable implementations of the software now so the past six weeks have been somewhat

uneventful which is good uh we've seen a number of unfortunate small slashing incidents due to misconfiguration and kind of over optimization of setups and we've seen i think client teams have continued to roll out updates focusing on stability optimization and some security improvements so nothing crazy on that front like all very expected um and now we're uh circling back at the new year and uh figuring out well we've known where we're going and now we're heading heading in that direction um and primarily that is uh you know fulfilling what the beacon chain is for uh the beacon chain what it is is uh we

the community bootstraps a new proof of state consensus mechanism that is intended to be the more sustainable more secure and more scalable home for ethereum in the long run and so to do that we have to ultimately extend the functionality what the beacon chain does today is come to consensus on itself but in the future the plan is for it to come to consensus on more interesting things for the community first and foremost what we consider ethereum today eth1 slotted into that consensus mechanism hot swapping uh the proof-of-work consensus mechanism for that of the beacon chain

um and for scalability gains um some sort of sharding to be able to handle more capacity in some sense uh with a single consensus mechanism kind of all unified in the same network so i i that was a lot uh but if you want to dig into any of that please yeah so there's something i want to jump ahead to because i um i mentioned ben edgington's uh article uh in the block it's not what we called it um i'm not terribly familiar with it but anytime ben writes something it's it's generally very high quality and you just touched on this hot swapping out the engine um the eth1 engine for the eth2 engine um

and i've i've created some rumors about test nets that um might be implementing this sooner than later can you tell us number one more of is that different than what we previously understood what happened and if it is different how and will it happen tomorrow so the fate of what we consider at the end today uh what is often called eth1 is to live in this new consensus mechanism um that has been the intention for quite a while i think early in the days it was uh contested as to exactly what the fate of what we call each one is but eth1 eth2 there's not really like a sequentiality there it's more like eath

one will live inside of you too the sustainable home and so the how exactly eath one lives inside of that consensus mechanism has evolved uh especially over the past six months as there's a pretty active working group that focuses on this merge uh this has members of what we would call like the two side and members of what we call the eth1 side because this is ultimately the unification of those two uh sub components of the protocol and the unification of the software that drives the components of that protocol their i would talk about this in a blog post soon but in light of a more like roll-up centric roadmap in which um

kind of ethwan is this like first-class citizen of state and execution and transactions and scalabilities through more sharded data environment rather than additional state and execution um it makes sense to bring elevate eth1 in this new context to be a first class citizen um and maybe not be this like disjoint shard and instead live adjunct inside of the beacon chain so like the beacon blocks we know and love today that have all those attestations and things actually have like the user layer payload which are transactions which is essentially the contents of ethernet blocks today um that's that is what is very actively being worked on and being vetted um i think it's a really beautiful simplification design it's funny because

it ends up putting us you know in early ffg proposals essentially we're going to bolt on this ffg thing to existing eth1 blocks in this path ultimately we created this ffg thing in a different context and now we're going to bolt the eth1 contents into them um and so it is it is a nice design it keeps uh eth1 and kind of all the value of ethereum like very close to the core of consensus um and it's it's very simplifying in the sense that uh with some quick and dirty engineering uh a couple people were able to like put together um like take an e2 client which would drive consensus take death and kind of pull out the proof-of-work guts

um and and and put them together in a way to give what are near like not production grade in the sense of specs existing but like production grade in the sense that they reuse e2 client which is production grade they're using these two client e1 client which direction grade kind of in this context so what we do have in this working group is um some local test nets uh that take teku which is one of the e2 clients to be the brains of the operation and take a modified geth to kind of be the heart uh to be that access into the user layer and we can run local testaments with it today uh there's some like refinement and stuff going on that and also a lot of write-ups so that we can help share our work and bring more

people into this uh this working group um but i don't know i don't know when they intend to make like a public version of that but i imagine there'll be something like a sandbag a sandbox you can uh play with relatively soon so the nice thing about this path in which one is more closely integrated into the beacon chain is that it's really not a um it's really not much complexity on top of what the beacon chain is today okay uh so i have a couple questions to follow up um let me ask this one first what have you observed this uh this testnet merged client and did you feel like it was it um was it robust enough that it looked

similar to what might be used or was it like a proof of concept um it's a proof of concept in the sense that like there's not a super concrete speck written down uh but it is robust in the sense that it looks very similar to what very likely we turn into specs um you know it's adding a couple it's adding a couple of fields to the beacon block um and it's adding a couple of fields to the the beacon chain state um and uh and then adding some callbacks to a local e1 client saying pretty much like is this payload this user layer payload which is ethereum transactions the stuff that we uh normally use to interact you know is it valid with respect to you know what how gather how on each one

client sees things um and that all of that is a very viable path and the way that these clients have kind of been conjoined in this proof of concept um is representative of i think how things might play out but would i go and run this tomorrow and say this is production grade no i mean first and foremost we need like lots of consensus tests and all the like all the goodies that we we normally want but is the proof of concept very important in that it one demonstrates that utilizing these two pieces of software in this way is very viable um and it in the event of say an emergency where we need an emergency fork from proof of

proof of stake uh we have like i think we could put together pretty quickly um you know we don't want to have to put together something like that very quickly uh but i think it's i think we're in the position where we understand the moving parts and and could could bang something out cool um you had you acknowledge that we've sort of flipped the priorities or or the where the load will be maintained and you know initially it seemed like eth1 traffic would be a second class citizen as a shard and now if that's moved on to the beacon chain does that make shard data a second class citizen or is it um it doesn't so shard data if there's no state and execution on it

um so they don't look like eth1 they look more just like layer one data that can be utilized by the rest of the system um it one it wouldn't make sense you couldn't uh bolt it directly onto the beacon chain uh because there'd be too much of it that's the kind of the point right um so in that sense it's not quite as close to the core consensus and instead committees kind of check it in and not everyone has their eye on it at the same time that's kind of one of the core tenets of charting but if they don't have state and execution there's like massive gains in consensus complexity both for the construction of the system from an engineering perspective but also for that of the end user um and so this data can be leveraged for scalability and types of things like rollups and other

other fun things that scale with layer one data in like a context that is more familiar to existing applications uh both from like the eth1 chain existing beacon chain and also how these rollups you know you could have a roll-up that threaded across three shards uh for a single threaded very similar to ethereum environment so you could use deploy existing applications in massive scale across these things so um not a second class citizen but by design not as tightly coupled into the into the beacon chain because you have to do that to get the scalability that you want yeah interesting great thanks uh i'm going to

oh let's one more thing so i think i don't know if you said it explicitly but it was a statement of maybe a shifting of priorities to maybe to merge versus sharding um i would say that a merge is further along than we expected at this time um and uh might viably be the first thing to do it before sharding but i i think in q1 a lot of what we'll be doing is writing writing specs writing tests vetting kind of the various paths at play before one is is concretely prioritized um to bring to production first i it wouldn't be right to put you on the spot to say this but

is it possible that we might begin rethinking the two-year commitment um to lock up um so to your commitment you know that's thrown around uh pretty much it's a commitment until there's liquidity uh at which point is likely the merge um if the merge is before sharding then i think you could see that earlier than two years if the merge is after the building out of sharding then i think the two-year mark is more uh what you'd expect so definitely there's like a lot of technical considerations in prioritizing this but there's also community considerations right like if everyone's like screw this we don't want proof work anymore like we have like let's bring this merge thing let's like unify the protocol and then extend it um then you know that's

that's a very valid thing for the community to demand and maybe maybe precisely what validators will demand i don't know that's that's when ethics 10k and everyone says i want my liquor i need out that's right how do i get out i didn't know all right uh somebody else want to uh to badger danny for a while yeah sure so um going along and going forwards are we gonna see the same like big push for public multi-client test nets for them absolutely i think one very important thing is that um creating and forking and creating new versions of test nets

is not nearly as difficult as it was a year ago right like to get clients just basically playing together was a feat in and of itself um but similar to kind of the eth one dev cycle uh you'll likely see some like transient short-lived test nets something into like the yolo stuff where there's not much of a commitment for long-term maintenance and then eventually when you uh are moving towards production you fork your test net first then you fork maintenance um something that i didn't mention is that uh in discussions right now is something of like a beacon chain like warm-up hard fork in mid-year uh this is being discussed on the specs repo it's been discussed somewhat in the eth r d discord and i think

uh over the coming weeks will be discussed more broadly um in additional context like on you know with the stake in community and that kind of stuff some of this is some some cleanups in how state and blocks and things are managed underneath the hood and some of it might be adding additional small features for example um adding the notion of a sync committee which uh is very similar to the kind of committees we have but it's a little bit longer lived and serves uh like allows like clients to to hook onto the beacon chain really easily um so like this is a relatively minor feature it's relatively high value um and it allows us to kind of go

through the fork mechanics in mid-2020 before we go through something a little more ambitious like a merge or adding sharding and so that is also on the roadmap and before you'd see production grade test sets of merger sharding you'd see this kind of stuff you know maybe in q2 as we begin to kind of position ourselves for maybe a summer fork of some of the smaller stuff and that would be that would require hopefully we'd see some community engagement on these fork test nets and certainly community engagement on mainnet people essentially running an updated version of their software that add a certain epoch uh you know extends and adds and slightly modifies some of the rules for the extended capabilities so yeah we

will see these cycles of test nets and and be engaged with the stake in community throughout so that people can kind of test things out and be involved i think that kind of leads to the next question which was like what can you stake the community do to help with phase one i think when we look back to phase zero the incentive to participate in the test nets was kind of get phase zero running on the mainnet and all the probes that we were trying to distribute but now that everyone has their setup um on the main nut is there anything that you can think of on on how we can incidentally incentivize people yeah offer more webs i mean

so some of this on um we might make a push to do a a large relatively large test net to test things out to test like the fork mechanics out uh we also might make a push to make sure everyone has clients updated and all that kind of stuff i mean but a lot of the a lot of the heavy lifting i think for the next few months is really on um r d and production engineering and kind of getting things to the place where we can put these upgraded clients into into your hands and so until then um i'd say be engaged as much as you want on technical proposals um but from there like keep keep mainnet alive while we go and do this other work um and again on merge on merge and

charting some of these larger upgrades that we expect to see reach test nets more and later this year um there's probably a lot of other ways to get involved like especially on the merge if we have a merge test net you can come in and like actually run transactions and like experiment with load and try to push the boundaries on on what it's doing it's worth pointing out that all four of the launch clients have as far as i know been hiccup-free um for the past six weeks which is i mean that's that's a great um testament to the work and success they put in there so um i i haven't said it at all but thank you to the implementers for uh for putting the work into making sure

that those work well for us so thanks for that yeah absolutely i mean uh the fact we have four robust clients is uh pretty fantastic um i know usage is pretty it varies across the clients i encourage you to use a minority client check it out but even just like to have failsafes in the event that something does go wrong with a one of the larger used clients there's some pretty awesome stuff there big success thank you i think now that we talked about poeps maybe patricia can introduce we completely forgot about it yeah but i also wanted to ask what does danny think about having a pop-up for a slashed people

because we refrain from doing it because we didn't want to incentivize bad behavior oh that's funny um oh man i actually felt like there would there are people who have enough eath that they would burn three or four eve just to have that valuable of a poet and sometimes that people is yourself right yeah no i don't not me i was referring to patricio worth halter that's what we call a perverse incentive um i don't know i mean i won't i'm i'm not going to get my i'm not going to get involved too deeply governance i think that's a great life

it's a pretty funny one yeah that's an issue because we want pop to be permissionless for anyone to be issue a token so far i'm guy keeping it but it's a challenge right what i would do if someone does it without my blessing but i say i i mean i'm i'm getting the hint that we better don't do it well and one way to do it and hopefully they're not listening right now is to retroactive you know do unit swap style don't say you're gonna do it and then like at the end of the year be like okay everyone got slashed in the first year but no more air dropped for everyone that got slashed i mean i tell you they are listening because this call has poab and it's uh yeah so that wasn't a commitment no one

no one called me on that we're gonna watch the flasher today and find out that uh god i mean but the theory of community loves burn deep so i've been distributing the power for this call using the zoom chat um if you didn't get it chime in somewhere so i sent it to you i'm not doing it as i was doing it on previous call 100 now you either have to have your video on or to say something meaningful on the chat or or present yourself in a way that you are different from a bot because some of our popular calls got attacked

and because the pops of these calls are going to give you benefits like discount on devcon or some other things that we are cooking if the con happens to happen uh we have to be quite uh strict about civil resistance so if you didn't get your pow up and you really want it uh just message me privately on zoom or discord or someone else and i will deliver and we have many more questions so don't waste our time on this i don't wanna waste too much time but i may have created a um a contradiction here because and i don't know uh michael have you have you entered that have you begun that process uh i haven't shared the link so whatever you want okay great then no worries we're fine um if if you would like to claim a poep for this uh you will need to join the zoom

if you're watching on youtube uh the zoom link is on reddit.com eatstaker it is the second pinned post and you are welcome to hop in and show your face to patricio um yeah just turn your video on everyone that has video i'm dropping it awesome and so those were the hey evan those were the questions that um sorry i just was just scrolling through and saw evan van ness um those are the questions that we had primarily as a community and we we did request uh submissions and we have some really good questions and what we did uh the e-stakers did was rank those um in terms of what we felt were the most

important to be asked um but i want to plug something before we move on and evan ben s maybe think about it in one of his comments um there's a huge opportunity so obviously for the sake of community to get involved like test nets are coming keep those clients stable and robust um but there are in terms of like getting involved in a more technical standpoint obviously like going and helping client developers that's huge building out tooling that's huge but one thing i'd really like to see more of in um this year is to expand the scope of like testing and diligence and things like that uh you know if we had five teams that showed up with like proposals to stress test test networks do adversarial

things uh do some additional fuzzing all that kind of stuff like these this would be no-brainers so like if you know of if you're interested in this kind of stuff or you know uh teams that might be interested in this kind of stuff please reach out and i will probably putting out like rfps for that kind of stuff soon but a great way to get involved is like to to battle test and harden this thing because uh we need to get it although it looks good today we want to be like a top-notch utmost assurance that uh this thing is ready to be our new home before the merge awesome thank you um all right community questions um michael would you like to get us started with those yeah so one of the

community questions we've got here i would love to see validator exits to smart contracts made possible specifically for projects like rocketpool to be truly non-custodial what do you think about the importance of non-custodial staking pools and the steps needed to be taken to make these pools possible yeah i mean first and foremost i agree um i wrote up a proposal that got a little bit of conversation i think before the new year and has been stick just waiting uh for probably me to pick it back up um and there's a proposal that allows us to do this relatively straightforwardly today without making any changes that you can change

essentially changing the way withdrawal credentials are there's a a couple of downsides to this that were discussed the primary being that [Music] such a even though you could have ultimately the custody of these funds be controlled by smart contract you can't exit these funds with a smart contract um and this is because e1 contracts can't talk to deakin chain um and we are not in the position to be able to design how that might work today um and so the there is still like a layer of trust you'd have to bring into uh for things like rocket pool and stuff ultimately you're assuming

this key will exit at some point or will transfer funds if transfers exist um and that's that's that's the most major drawback for this um i think that we're going to we'll pick up the conversation this week um and look to make a decision like relatively soon um the decision i think there's like pretty high demand i think i need i needed to circle around with people like rocket pool and stuff and like talk about this particular sticking point and see if it is uh in their in their designs um the thing is you conversion these withdrawal credentials and create any we can create other ones in the futures but this this one would just have that particular drawback um and so that that's that's the blocker but uh it it's we need to pick up the

conversation again um now that the new year we're in the new year so yes but the answer to the first question i think this stuff like could be a huge boon to the diversity of the ecosystem um you know i'm super pro hobbyist taking but these more like decentralized um pooled and institutional players i think would be really valuable to get on main net especially as like the centralization on staking providers continues to happen and as more state comes online absolutely thank you okay so as we add functionality to the eth2 chain in the future phases should stakers plan for additional hardware requirements for either the beacon or validators right so the hardware requirements of

the merge would be essentially that of today you would need an eth1 client and an e2 client um if statelessness happens on um for the eth1 chain which is in the works i don't know when exactly uh that would make it to mainnet uh then you could potentially reduce those uh requirements as like you need to be able to process each one stuff but you wouldn't necessarily need to be able to um have the entire state with you which uh so you could see reduction requirements that over the next couple years depending on how that goes beyond that with sharding um the whole point one of the one of the points of sharding is that

you can participate in consensus with consumer machine that is you can run the beacon chain um and a small number of shards you know call it one two three maybe ten depending on what and what your bandwidth is and distance stuff but one of the points of sharding is also that the system in aggregate handles more than a consumer machine can handle and so if you're running a single validator uh you know i should be able to run which would have the requirements of just a small amount of duties to the shard i'd be able to run on my laptop no problem probably be able to run 10 of them no problem once i start running 64 1000 something like that uh i'm definitely i need i'm gonna need

like pretty significant uh requirements and this is gonna manifest both in terms of bandwidth and in terms of storage requirements because there's storage requirements to to the shards and so yes things will in a sharded context there will be additional requirements on validators and this is going to come again in the form primarily of uh additional local storage uh which you could probably expand on your machine pretty easily and some additional bandwidth requirements which uh for home staking depending on how many validators you're running could be difficult you know if you're running a hobbyist amount of validators at home i think you're gonna be fine if you're running a thousand it's it you know might be tough but if you can uh get that bandwidth going

then it's it's probably worth it too so uh this is something is that right the hobbyist validator is about a thousand pounds right yeah yeah something like that um so so the whole point is again the more validators you're running up to a certain extent is going to scale the amount of requirements you're running so if you're running like a megaton amount of validators and getting by today you're going to have to start thinking about what your setup's going to look like in the future that's also an argument for the merge before starting because if you can exit and you can have liquidity on your funds if the technical requirements came unexpectedly uh you could you could exit and kind of rethink your whole

setup whereas if we do sharding first it kind of puts this requirement on validators uh when they don't really have much choice in whether they continue to operate so that's another argument there now one of the questions i personally had was like assuming that um sharding is live and does aerial validator only choose to be on one chart or how does it look like so for instance if an explorer wants to display all the charts um what kind of um syncing type will there be like and is it going to be like an archive type or how do i get the data from all shorts right

so as a validator specifically you're constantly like assigned to a shard to be doing doing work if you're running 10 validators you can be assigned to 10 shards be doing work although some of that work is sometimes segmented so in the uh you could potentially not have to take on the entire burden of that entire shard but maybe more like one slot per epoch of that sugar so that's how the kind of the technical requirements happen in terms of a user um a user being you know some casual user that wants to peer into a shard there should be configuration on your local node like i want to follow shard one i want to follow shard two i don't wanna follow these other ones um in terms of a block explorer they probably say i wanna follow all of them uh which ends up being if used at max load pretty

high requirements and then exactly how the apis evolve and how this stuff is exposed to end users is definitely tbd one interesting thing in the current charted road map is this is just sharded layer one data there is no canonical interpretation of this data in terms of layer ones there's no like transactions in them there's no well you can put whatever you want but the interpretation of it from like a layer one block explorer might be nil whereas but from a layer two perspective some uh you know some roll-up might use all of the shard blocks in shard one for like their roll-up chain and so as a as a block explorer i might choose

to do like uh you know a native integration with that and like show it as a first-class citizen like what the actual payload of those blocks are so there's a lot of like weird design things to think through especially on the block explorer front um but i some of that well i know some of that actually is stuff that i think block explorers are going to have to tackle sooner or later or even before shuttered data because a lot of different roll-ups are coming on coming online and assets are kind of assets and contracts are beginning to be on different layers and so i i imagine if i'm not intimately aware of i imagine block explorers are going to

start trying to figure out how to elevate some of these l2 integrations into first-class citizens um our friend jc asked in the chat um how is the community being polled about the priority between the merge and charting well this is one of the first times to talk about it publicly um so it's not being pulled right now so there's two things at play one is the technical requirements of each um if one where over the next few months became abundantly obvious that it was ready to be executed and we could could be rolled out um i think it would not be much of a conversation because i think we just want to execute

uh not wait while the other thing is kind of being solidified if both are like very much a viable strategy immediately viable strategy and specs are ready and things are ready to go um that is it's a harder conversation i mean it's a broader conversation and uh jc i don't i don't know um but it's something we're gonna be able to talk about in the next few weeks next few months but um i don't know i mean i i i probably i have i'm beginning to have a bias uh and i think i've stated that somewhat clearly here um i'm excited about an early merge uh but you know i think i think don't know maybe some poets we can do some poet

votes okay um yeah what should be the framing of the question because if you ask people what do they want they want everything all right so if charted data here's one way to think about it if sharded data is to give us and this is again my bias coming through because i'm framing it for you that way if charted data is to give us scalability through rollups um but optimistic roll-ups that are coming online in the next very soon already give us 10 to 100 scalability in those contexts what's another 100 from sharded data in the next like year or so and so from that standpoint i'd say letter two max out ethereum today and at

the same time get the merge done uh and then extend the layer two capability through additional sharding uh but how to frame the question um i think well for one i think the question would be better framed once we've done a little bit more technical due diligence um and then from there yeah i don't know it's it's it's tough because it's like technical de-risking it's like encumbering the e2 development process with all of one stuff it's like uh there's a lot of different pieces and so it's probably each person's gonna have a different set of priorities um but when you're framing that poet question you can knock on my door and i'll try to help okay we'll do that so is it the same group of people

working on both sharding and the merge or is it two groups and you know you can't really prioritize it um right now both of these are more of r d efforts rather than production engineering efforts um in the sense that uh we have this merge working group that's been working on r d for the past four or five months maybe probably more maybe six months um and we have this data availability shorter data availability and spec writing that's primarily being done by my team or the team i'm on at the ef um what we're doing right now is working on communication education with the client teams who will in some capacity be doing something like

the r d here and they'll take some of the ideas we have work on some sample integrations in their clients and then kind of begin the feedback loop into the spec writing that on both can on merge and charting likely can happen a little bit in parallel it's really when uh because you can only throw so many resources at an r d project it's really when that stuff shifts from r d and spec writing into production engineering that i don't think you can really like do both fully blown at the same time like you you then have losses because you're trying to build two production versions of different production features you're trying to ship two different test nets you're trying to do you know so at a certain point you know in the next handful of months um i think some sort

of decision needs to be made so that we can uh kind of put the pedal metal in full blown production engineering but right now yes things are split and that's i think a good thing but sooner or later they won't but won't be able to be fully split so is it fair to assume that and so i've missed the past couple of implementers call and i'm sure this is answered is it fair to assume that the implementers are currently polishing the beacon chain clients and not doing new implementation is that right yeah as of now um i think what you've seen since december 1st is a lot of polishing um and only in the past week am i having these kind of like conversations with different engineering

teams to um they've been heads down right they've been focused very very much on phase zero delivery so there's a lot of kind of like educational education initial conversations get people up to speed to better understand this stuff um and definitely one of the topics of the calls from kind of here forward is the the balancing of like keeping stuff production stuff moving and kind of filling in the gaps on spec writing and yeah r d here well that really makes sense i mean i i didn't zoom out out on it until just now but to realize that um just like i've been totally focused on the beacon chain launch the implementers have also been solely focused on

phase zero and it's it's not they need time to get up to speed on the future just like the rest of us do so that's something that yeah absolutely and some you know i i see terrance sitting right there uh he's he he's one of the engineers that kind of constantly has his eyes peeking at uh different r d and stuff but a lot of engineers for very good reason you know have just been like heads down shipping this um so there's there's uh definitely over the next few weeks a lot of like information exchange back and forth and bringing people up to speed so we can put the pedal in the middle one of the things um i have been pushing for

in the community is is a single click installer for all of the different clients um and and i i don't really know the right direction to push i just sort of yellow that out on you on twitter every now and again and see what happens um is that is that dream of a community sponsored open source single click installer for all four clients is that something that has legs or am i just dreaming and i should dream elsewhere i think that would be a very reasonable product or test and there's something uh i think geth maintains a product called

puppies which helps you like spin up test nuts and different clients and different things and it's a command line tool but it does try to make the management of different clients in different contexts more generic and so i think it's one i think it's pretty achievable and i think there's i think there's high value there one to kind of mask some of the complexity to some of the end users and even make certain things easier like safely uh switching client types you know by default uh when i change from client a to client b it moves my slashing you know slashing protection database and things like that so uh you know i think if the right team were interested that it could be a like

a pretty fruitful project um it's you know it's probably not something that certain users would use like i would probably end up just hanging out on the command line because i want to uh but i i do think that there's likely some value there do you this is another like you haven't prepared for it and you probably haven't thought about it would that installer look better as kind of a ppa apt-get or would it look better as a docker kind of um which which would be the right direction if someone were interested in working on that yeah i don't know maybe you don't know and that's okay because i i i don't know i can't i i don't necessarily i'm i have trouble um

assessing that type of user's needs and better understanding what they need um and you know and i think somebody else is better suited to do that um i'm way too deep in the guts of this thing to like i think you can't imagine being someone you're not no no no i mean i could i could try i could try but i think there's probably people much better suited to um to figure out what that that ux need is sure yeah one of the questions was what are your goals for eve to development in the first half of 21 yeah earlier but we've covered a lot of that one would be to

uh concretely get this like hardport one moving um all that that stuff we're fine in terms of the specs uh to bring both merge and charting from uh kind of like iron out the kinks in the r d process uh do some like proof of concept testing and networks and stuff and bring that into concrete specs um and to position engineering teams to drive really hard on one of those two mega features to deliver production grade test nuts of it by the end of 2021 to be able to bring that stuff into reality next year

i kind of like the next question um are there any plans to define a common beacon chain api meaning that one could run a lighthouse validator with a prism beacon node so there is a repo called e2o apis um this does define a standard suite of apis that are first and foremost focused on like kind of information accessors uh but they also do have a namespace called slash validator that does define the minimum communications between a beacon node and a validator client i don't know the state of conformance to this api across clients i think a few or at least like nearly

conform it um and so you could i think i think uh lighthouse and tekku are probably the most conformant right now so somebody could try running a test uh between those validator clients and nodes um i don't know if anyone's actively attempting to do that there's i see the value there um to be able to like quickly swap your validator client with a different node um is is very appealing um but there's also this like slashing protection database so you could swap your whole stack for a different node uh safely actually i remember a lot of these client a lot of validator clients have like the backup beacon node api now where you can specify multiple beacon nodes in case one goes down i

believe someone tested either taking i think the lighthouse about her client and pointing it to um both lighthouse and tekku and there was a small misconformance in the api so that actually there are a couple and and it was reported so i think uh it might be fixed now so it's feasible and potentially possible today that would be an awesome like three minute clip to watch um just like you know oh this is something proto would do because he has like eight um he doesn't use a screen what's the alternative to screen that you use uh somebody help me anyway so eight screen sessions where it shows you know running running and then it shows

where the validator the beacon chain stops and it kicks over the other one that would be um i would i would heart that on twitter so whoever wants to put that up great i think i mean the next question i had was i think more towards tehrance and he will probably answer won't know it but is there are there any plans to just support the easter.api instead of continuously uh managing the prism api and lighthouse api what was the question sorry oh terence there yeah we definitely have plans to support the api it's a work in progress right now and uh we're aiming to get out in like

q1 okay that'll be a huge win so we have been through most of the most of the community questions and i know um i i i'm wildly impressed with the information we have but i'm sure there's still more that we can learn now is a great time to mention that we are offering a poep for this call and the way to get that poep is to show your face on zoom to patricio worth alter and then he will dm you with a um a claim code we are actively actively working to thwart our farmers because we do intend

in the future for these poeps to have lamborghini style value i just threw that word out on accident yeah so i i've sent almost um all that's been requested um including like 20 on discord um if you were participating on our discord and you reclaim a pop-up um and you didn't have a camera you still get it uh if you didn't get it because you look you do look like a bot like if you never interacted with our discord you don't show your camera your nickname looks like your catwalk on the keyboard i'm sorry to say we are gay keeping this one

because we really want to have humans on the call but i think most of you are getting them so so um danny did you get one i haven't i haven't either so we we must be i delivered both of you because you were on top of the memories because your hosts and and even if if by the time the call finishes somehow i glitched and you didn't get it just keep a screenshot that you were on the call and something that proves that you are a human like like if you have a pop collection that has a pop of an event you mean in person like defcon or any other that's all we need uh the only thing we want to prevent

it's 20 people that have been likely managed from a single location with a common center software and and that's what we are not giving but we will come up with something in the future we just want to put a very strong captcha because that would be annoying for most of the other users and and i guess we can go back to questions i don't know if if taker has any i am a little bit ready to open this up to the community um which um will need to be a little on guard because i see several people we don't recognize um but um yeah is everyone okay with it let's open mic uh those that don't have

cameras maybe they have microphones and they can stop whining privately to me okay should be of that mute um let the texan talk we have a texan he's also a mate drinker oh i see now say something evan okay well um i am happy to call it a call let me

think real quick if i have anything else that i'm going to give a quick tldr yeah we bootstrapped ethereum's new proof of stake consensus it comes to consensus on itself people say the beacon chain doesn't do anything that's precisely the point it's consensus uh before we uh add the like 100 billion or whatever the hell theorem chain into it uh we want to kind of vet it in production so that's what's happening right now um but what do we want to consensus on when it comes to consensus on ethereum what we normally call eth1 i want to come to consensus on uh things that will help us with scalability sharding charted data um that's what we're working on in 2021 if you're a validator and obviously keep your running keep

things production uh stable and keep your eyes peeled for some of the fun upgrades and test nets throughout the year is it fair that that one of the big changes we're looking at is that these things are which were previously called phases are now less sequential and more parallel yeah i i probably haven't used the word phase at all on this call um i'm i think that it's you know my framing is we have a consensus mechanism we want to come to consensus on some things um and those major features are emerge and sharding uh they are being very actively worked on in parallel um and we'll go out to main that when they're ready um you know phase one phase one point five phase two like uh there's certain like

expectations and kind of implied sequentiality with those uh that don't really hold true uh so instead uh you know shifting the framing a little bit and then once you have the spec it'll be more clear whether merge or sharding will come first yeah um it's a little bit of the spec and it's more um kind of r d betting some ideas uh running simulations and stuff and and better understanding the complexity of each task as as you know not that we start the price but the price is going up and that that does tend to bring more interest to ethereum and i i do believe that it's likely a lot of new technical eyes will view a

video like this and so if if someone is rather new to ethereum and has a developer background and wants to contribute to eth2 specifically what would they do to get up to speed and get involved i mean in terms of getting involved everything is open source uh and i know it's easier said than done sometimes but contribute early contribute often like there are fix a typo in a readme um you know when you run into a function call and like a tool you're using and it doesn't it doesn't work properly or the documentation's bad just go in and like file an issue or fix it um so like getting involved once you

once you get involved just a little bit it's it's um infectious like you can't you can't stop um and so uh get get involved early get involved often um and then in terms of getting up to speed uh there's more of the kind of like what's going on pieces in terms of blogs i think ben edgington's uh what's new in e2 piece is very good um i try to do some like shorter um updates called e2 quick update you can check that out and then uh more broadly i would pull down some software run it uh see how it goes that that's a big learning experience and then there's these like annotated specs one maintained by vitalik and one maintained by ben

edgington that if you want to do a deeper dive to understand like the actual mechanics of the things going on uh those those help you bring you up to speed um there's also the eth r d discord uh it can be a little bit intimidating probably at first but just join work uh pick a channel that is uh in a domain that you're interested in and just kind of see what's going on check out the chat or eventually you'll have something uh to contribute and say yeah i think you'd also plug each.research or ethe r-e-s-e-a-r-dot yeah yeah if you're more on like the kind of theoretical side um each research is a great like kind of long longer format forum uh where people exchange ideas and debates and stuff if this might be a stretch but if if you

were going to suggest a place like a first pr repo um is there something like that that's maybe low-hanging fruit that could use some uh um sometimes stuff at the edges of the stack my first pr repo was web3.pi uh many years ago you know it's like it's a wrapper around apis and stuff it's not super complicated but really if you if you're if you have a background networking then go look at clients go look at their networking portions of their stack if you like messaging with json and apis then check out uh lodestar check out you know web3pi now has the wrapper for for the beacon chain um and if you like getting your hands really dirty then

open up a client and see what's going on uh sometimes teams mark things as good good first issue and you know there's even if e2 clients aren't your thing uh the stack is huge there's a lot going on there's tons of teams and people are generally like very very welcoming and receptive to helping hands you mentioned lodestar and i haven't thought of uh chainsafe in a moment um chainstay floatstar uh are are they are they focusing on a different direction now and or are we still maybe so the the big value from a lodestar implementation always has been in my opinion the um javascript libraries that come out of

it and kind of enhancing the developer tooling additionally uh being native javascript there might be some like good browser integrations for light clients and that kind of stuff and they've also um at least early on when we thought these were a little bit closer they were doing a lot of like light client r d uh they also that team helps drive the the api standardization process and so i think that they are following the head of the beacon chain and kind of an alpha capacity and might very well forever be in kind of an alpha capacity in terms of running a mainnet client but all of the things that come out of the creation of that client is a big boon to the developer ecosystem and the browser ecosystem for

free too um so uh you know will you ever be running validators on lodestar i don't know maybe but we'll accept it but but will libraries that are a product of lodestar's creation uh be under the hood of something you're using almost certainly at some point especially if you're interacting with browser stuff does it seem to you right now that that the four uh launch clients so at least from my perspective those four seem like they have legs and they will run um do you anticipate without i don't want to put you on the spot any changes in that number um um it's a good number

uh these clients are strong they're all very committed to the next couple year process so i don't expect it to go down um i know that nethermine had like an early e2 implementation called cortex although they're focusing on a lot of their eth1 stuff which that component of the software i think will be intimately important in the stack after emerge so you know they might pick up the cortex work or not they might focus more on that user layer um and i've i've heard rumors of uh you know other secret clients but you know in the past i think we've probably had 102 clients maybe 11 22 clients in total um and so you you might you might see someone pop up

we'll see so that's a metrics clients is there any and maybe proto might be able to to poke on this as well uh or some of you you people who are deeper in the data do we have any inkling or evidence that there's a difference in the rewards per client or is that something that's still nebulous um in terms of rewards not necessarily we likely there are some clients that are more performant than others at the time depending on the specs you give them so if client a uh you know which is not the case is twice as as cpu hungry then you might

need to give it more resources to get that profitability um but so ultimately if you have to have a more expensive sheen to run something then like you could take that that out of your margin in terms of profitability so there are ways to kind of compare but there's not great comparisons now i do know that all clients do run at least some validators and i do know that all those validators uh run exceptionally well on the infrastructure that the client teams are providing and so i know that all clients can perform well i think one of the hot items right now is that like random three percent drop we see in um attestation correctness it's actually the ffg target

correctness that we see if you're watching a block explorer this has to do with and i've been talking about this on the ether r d it has to do with one a minor fix to the fork choice that needs to go into play and slowness in either the production or the gossiping of epoch boundary blocks sometimes um both so the first will be fixed in kind of a spec fix and modification which expect to be rolled out in the next couple of months um and i think we'd greatly see a reduction in that issue um and the but the latter issue um there is some investigative work uh to figure out which clients are kind of having this

hiccup at the at the epic epoch boundary and where exactly is it in block reduction or is in in the way that uh this block's being gossiped um but so whether that be a client two or all of them uh there's a little bit of like optimization work to do there that might marginally affect profitability awesome uh so it sounds like the next time we'd want to check in with you is at this um hard fork that's you froze superfiz i do that every time that's my trick um it sounds like the next time we want to check in with you i'm gonna have to change this box because i've tried everything with it um the next time we want to check in

with you is the hard fork that's coming around mid-year does that sound about right uh yeah i mean that'd be a reasonable time i'm always happy to chat um i think definitely over the next couple of months there'll be a lot of conversation around that uh as specs are being refined and kind of the the features are being selected that go into that is it too optimistic to um to believe that one of that that either the merge or yeah that's six months away i was gonna say it's too optimistic to believe that either the merge or the sharding could kick in at that six month and even i realized that's waiting no and and one of the reasons that we want to do this fork mid-year is

uh as as kind of a warm-up um nothing crazy is going to be in this fork a couple of minor features a couple minor refinements um and it really is like testing the coordination of getting one of these things out uh before we do a more major feature um but i do expect to see some movement and fun on both those things this year yeah great all right uh does anyone have anything else they'd like to add before we wrap it up uh so this has been a wildly successful call uh i think i saw about 85 participants um and um i it's so awesome to catch up like i

i now feel like everything i've missed for slacking playing dark force over the past month like i'm i'm ready to get back in the game now so uh i i really appreciate all of your time danny yeah thanks for having me will you shield your the meme competition and our future yeah right so in order to get more less technical people in the community we decided to run a meme competition um to basically get people from e-finance and eve trade and all the other communities involved it's kind of a bait it worked a little bit let's see if you can get more people yeah so if you want to participate just

check out the reddit i will pin the fret again and if you participate you will also get a poep and some other rewards that are secret so broadly creating memes related to e2 staking or if one just etherium names cool i'm i'm very proud of my submission but i entered my address as the contract address i might go edit that now because i i didn't know there was going to be a poem some other things that we love to talk about we have uh ethereum studymaster.com uh which is up and running on vedica or michael runs that that's a great way if you're starting from uh early on to get into e2

read some articles take some quizzes and earn a poep for participating we also have um what are those groups called um collaboration groups yeah so if if you want to go to our new and improved website eatstaker.cc you can find a link to our collaboration groups you can sign up there and you'll be joined up with about four other people you guys will be in a private discord room where you can get to know each other and support each other and learn together i'm in love with that collaboration group program um we're going to be scheduling a new game night coming up soon um uh michael and i will be talking at east chicago uh january 14th um

and uh that's we we've got a lot of exciting events planned for this year so um i'm glad you all joined us and um we'll call it a day thanks again danny sweet yeah thanks for having me happy's taking everyone bye-bye and so i think i got most of the pop delivered because i delivered over