From Concept to Creation The Evolution of Lido Community Staking Module | SheFi Summit Bangkok

Transcript

[Music] hello uh my name is Sasha I am a contributor to the LI Dao I'm a product manager and me and my team recently launched the product and for me personally it was the very first product um I built from groundup and that was quite a journey I have to say and today I want to just walk you through the journey together with me not the entire Journey but just a small piece from like an explorational phase to understanding the problem and to even considering the design space and trying to choose the

right design solution all right so yes before we got to the design in the product we need to understand what Li is and what it does um okay and for this we need to uh dig a little bit into ethereum we all know that ethereum needs validators um to perform their duties so that our transactions can pass all right so uh yes the beauty of ethereum is that anyone can create a validator uh for this um you need a hardware a stable internet connection a some pieces of software and no more or less 32 is all right and this 32 is serve as a

guarantee that you will perform your duties not you but your validator will perform your duties on the network uh correctly according to the rules all right this requirement of 32 is creates one of the most challenging um issues on theum network that yes there are a lot of people who has uh desire who are willing to run a validator and there are a lot of people who has this Capital who have this Capital but the intersection between these two is quite small all right let's take a look how uh lier in particular and in general liquid St protocols solve this issue all right we

have this um is uh holders uh people who are ready to provide their Capital they put their is into the liquid staken protocol that allocates this is to validators validators perform their duties on the ethereum network they get rewards that are splitted between validators liquid staking protocols and uh is holders and it's seems like the challenge is solved but not really so the deeper we will dig the more challenges we will find now let's take a look at the validators side of things so as we already discussed the validator is a piece of a software that is set up um on some kind of hardware and no operator

is an entity or individual meaning that it is uh like a person or a group of people people all right so know the perator who they can be or who are they so I I would say that generally there is a division uh and we can say that there are solar sters and professional noors all right solar sters is just individuals like you or me or anyone else in this room who voluntarily want to run a validator their home maybe not necessarily at their home but um they usually combine this duty of maintaining the validator with their regular job in a real world so and it means that the um they have less time for maintenance and

it can affect their uh performance but the best uh thing about soloers is that that they are individual they independent from each other uh in terms of their like jurisdiction in terms of their dication in terms of their even like views and interest so that is exactly what eum wants to have uh to stay decentralized and diverse and the other side professional node operators these are uh organizations who do this as a their basically business so that's the way theerum validation validation for them is the way to earn money all right they are quite experienced they can um ensure the stable maintenance of

their validators but they are usually called a single point of failure and what I mean mean by that is uh this so if we take a look at this example we have 30 validators here all of them are belong to 30 different node operators and we have 30 validators here and all of them belong to only one professional node operator and if we imagine that something happened in a particular region in a particular City where One Professional node operator and one solar operator live what will happen we'll have this kind of picture with the outage uh so of course this is an extreme example and everything is not that obvious and professional node

operators do care about like diversification of their servers of their clients but it's just a good example to show you what I mean by this single point of failure all right uh how do no operators join um again two polar opposite approaches the first one is permission meaning that an another prator need permission from someone or something in order to join the set and the other one is permissionless obviously when uh you as an a operator do not need any permission to join you just go and join all right so now let's take a look at um the situation that lied ahead when we started working on

this product so there was just professional no operators that were going through the permission and boarding rounds is that bad uh yes and no so the this model um allowed uh the protocol not to set to scale quite quickly and to fulfill this demand in terms of inflows um of stake uh but again it was not quite enough in terms of the decentralization and it required some kind of action from the protocol to change this situation all right at this point all of you have the same context and the same knowledge that I had when I started working on this product so we are kind Alliance right now now let's go

into what we're going to do about this so we've identified the the problem that the the the the problem was that the light set was permissioned and was allow was allowing to join only professional not the periodist and it can't uh give you the enough the enough level of diversity and decentralization I guess and the most obvious solution in this case seems to be just okay what if we add a subset of node operators and for this subset we'll set up a permissionless entrance and that would mean that anyone can join there wouldn't be any problems with like operational stuff around like on boarding each and

every single individual node operator all right seems like an solution so but for me personally the most uh like the B the the worst news was that having permissionless entry to the module does not guarantee that individuals will actually join that was quite a surprise for me honestly so and now we as a big team who are about to solve this problem are going to figure out why it is so so there are a lot of uh reasons why they will not join starting from like social reasons that they just don't want or they think they're not aligned with

slider or they do not know about lier but we will consider just the very simple case they there might be a case that they just cannot join why so what if we just create a permissionless module with permissionless entrance what's going to happen will uh solar operators come to us uh they can but the probability of big node operators coming in and occupying the entire capacity of the module is way higher so we need to kind of prevent these professional node operators uh from occupying the module all right let's try the following we can

list all the addresses of professional node operators we know and just say no you are not allowed so this is for small players this is for soloers do not come here so we just reject the entrance for them okay that's possible but what we can end up with is that these node operators professional node operators will just create create new address that are not associated with their entities and basically that's what we call sibling and that's definitely not something that will help us to decentralize the validator set all right let's try to do the opposite what if we list all the solar node operators addresses and if we and allow only them to join so opposite all right that's

possible but the problem here is that it is quite hard to identify if it is a no perator or not because like on the um blockchain and on the etherum network you can like find any information about uh this address except for its owner unless it was like voluntary provided by the owner of this address so it is quite challenging to get this information uh and moreover and the biggest problem here is that if you list all of the addresses of existing solar sters it will mean that you are not allowing newcomers to join so and this issue doesn't solve uh in this approach all right so we tried three different approaches from the design

space now let's consider the following what if we indeed create this list of solar stakers and allow them to join but we will do it before we open module up for everyone so before we were trying to get two modules the permission module and permissionless module and we were trying to to split the this professional and Sol node operators between these two modules but what if we will not try to split them into two modules but instead we will split them like in terms of time so it means that we create a buffer for the module till uh the buffer it's a period of time during which the module

will be permissioned and we will indeed allow only identified solar stakers to J even though I told you that it is quite challenging to identify these addresses it is possible to some extent using the machine learing models uh that analyzes the behavior of this addresses so it is possible to some extent you can use multiple um sources for this but the idea is that we do not split them into like different modules because this permissionless module will be occupied by the professional noer but not from the day one so there is solution and we've just came to the uh like one of the design solution from out of many and

basically this is the same solution we made for a real product and I haven't named it yet but it's time to say hi to him so the product me and my team were building uh is called the community St module the first ever permissionless module for the lighter protocol that is currently live or main net in this early adoption phase so this exact period while which we're allowing only identified solers to join and we expect this module to become fully permissionless at the beginning of the next year so for those who um who've been involved in theum validation or agnosis

validation uh I definitely encourage you to check out this QR code and to check if you are eligible for this um early adoption mode for this early adoption period and um for those who are just interested um how this uh story uh finished with all the design considerations around CSM it was just one out of I don't know how many thousand around yes so yes uh check out the architecture of the module and hit me on telegram if you have any questions thank you for your 15 minutes that you uh the 15 minutes of your life that you uh give to me I want to thank my team for working together for building this

protocol and for you all of you guys from this audience for allowing me to let something that is very important to me into your hearts and that's pretty much it thank you