Not all L2s are built the same Understanding L2BEAT L2 assessment by Krzysztof Urbański kaereste
ETHDubai 2024 · 21 Apr 2024 · 23:11
Krzysztof Urbański explains how to use L2BEAT to assess layer 2 systems. He compares the properties and trade-offs behind different scaling designs.
Speakers
Topics
From ETHDubai 2024, hosted by DeFi Dubai.
Transcript
R okay hello everyone uh hello Dubai my name is Chris I I'm from El Tobit if you don't know what El Tobit is and what we are doing you will learn in a minute I hope you are having a wonderful day in this sunny Dubai and this is a great opportunity to talk about the end of the world let's talk about the end of the world uh who of you sold like played the game or or sold the series Fallout who played Fallout yes yes we've got some Fallout funds on the at the audience so like I don't know if you if you already watched the series in my
opinion the series are great uh I watched it when I you know traveled here um in the plane but one thing one thing didn't click for me in this whole lore why the hell did they they don't have internet and why why the hell they don't have crypto like you know they have computers right they have radio if they have computers and even Network computers and they have radio obviously they would have had internet and if they had internet I'm pretty sure that they would have had ethereum because one thing that I'm pretty certain about the end of the world that ethereum will survive like etherum is there to stay from all the crypto projects like I
personally believe that this is the one that is most probable to to survive any difficulties because it's so popular it's so used you know all over the world uh obviously I'm I have no idea about the prices but the network the technology is there to stay but the ethereum is not just ethereum right four years ago vitalic uh already drafted the the road map for eum to be the rollup Centric uh EC system so when we are speaking about etherum you are not talking about etherum itself we are talking about etherum and all its l2s that actually create this whole
ecosystem and the question is if itum survives would those l2s survive and that would be something that I would not be so certain about and right now we can already see that this there is this huge amount of value already locked in l2s like this is already we we are already living in the L2 World there is a lot of value but there's also a lot of activity the activity on L2 already surpassed by far uh activity on the main net itself so you know will will they survive will the L2 survive and one thing that I know
about startups and one thing that everybody should know about startups is that most startups fail and crypto startups are no different there is no reason to believe why crypto startups should be different than regular startups so we should all expect that a lot of those rollups that we have right now a lot of those l2s a lot of those projects will simply fail and they will cease to exist few years from now who who of you rem Solana sorry not Solana EOS for example who remembers EOS like yeah there is like one or two per two two people at the audience so basically many projects that we leave that we know today will also cease to
exist and the question is what is going to happen with the assets that we have in those l2s and there are two ways like there are two ways of thinking about this so we can think about the assets that we Bridge or mint on L2 as casino chips if you go to the Casino you hand them dollars and then he they hand you back chips you go you have some fun you play some poker play some Blackjack play some play some roulette and hopefully if you are lucky by the end of the night you have you know a ton of chips like a huge mountain of chips but let's say that you didn't exchange it for dollars
back for dollars but you simply to took it to your hotel room and then forgot about it you know packed it into the suitcase and left and you left for a year and next year you're at the same Casino you want to go uh exchange those chips but you see that the casino is no longer there it doesn't exist like you know it bankrupted so what are your chips worth right now and the answer is probably nothing and that's something that we would like to avoid in rollups and there is another model the banking model something that we like didn't like from of banks something that that the reason why Bitcoin actually um was created was that our governments are bailing out the binds like the one
thing that we want to ensure in the financial system that if you put your money into the bank you will always be able to receive it no matter if this particular bank will you know continue existing or not this is something that that we guarantee as a society as a you know in in in in advanced societies so we would like our layer to to be the same we would like our layer to to also secure be secured by eum there is this meme already that is you know that that layer TOS are secured by etherum but really are they what exactly what security exactly does itum provide and the answer depends
on what L2 are we speaking of of of how this particular L2 has been built because if your L2 looks more like this then I would say that it's pretty okay I will explain soon but if your L2 looks like this I wouldn't keep my money there basically I would be rather cautious you know it's okay to experiment it's okay to play with such an L2 but I wouldn't trust my money to be sitting there why that's exactly what we do at L2 be we actually analyze those systems and check whether the funds that you are putting in them are actually secured by etherum
that is lb.com website and what you can find on this website this is our risk analysis you can find all the l2s that we assessed and it's basically we try to have all the ethereum based real l2s on our website you can find like different risk vectors analyze there and you as you can see there are a lot of Reds here which means that we are no longer there yet not not yet there so first thing that we analyze is the what happens if there there is a proposal failure what happens if this mechanism that puts your data to the L1 to the at your main net what happens if this thing stops working
and unfortunately as you can see here like usually you have you have problem uh what would be perfect if you could for example self propos post the other the other solution that exists that is basically kind of uh lower on this website is to be able to use escape hatch to to somehow retrieve our funds the other thing like but but if there is an option like for example in arbit room we are able to self- propose but in order to be able to self- propose something the construction that we are using needs to allow for self sequencing right we need to be able to uh improve the CH like to push push
forward the chain to be able to exit with our transaction if you want to do anything with that and fortunately most of those Solutions already address this issue somehow if they don't I would be cautious because that would that would make it difficult for you to retrieve the funds then in order to be able to propose something in order to be able to advance the state of this L2 first you need to be able to reconstruct the state you need to be able to to retrieve the data that sits behind the state of this L2 and there is a question where does this data sit where can you where can can you access it and actually this recently became much more complicated than than than it used to be before
because of all those different data availability uh Solutions and the change in how we how we approach data availability but basically if your data is not available on chain or in some convenience way if it's externally um accessed then there is additional risk that you have to take into account and there is additional thing that you have to monitor and check to be sure that your funds are safe another thing which already touches uh more um the security of your funds uh while you're at this chain is how is the state validated and is there any is there any mechanism to
ensure that state is correct that EUR funds haven't been stolen uh in the past and as you can see like many many rollups not yet do not support uh any solutions here yet uh especially if this is optimistic rollup and it doesn't support fraud proofs you should be really cautious because you are actually trusting the sequence side that it um does the right thing then oh I missed one slide there should be oh sorry the something that I wanted to say here is that if you have the fraud proof there is this requirement that that Frost Proof has to be actually running uh long when when
you know when somebody is watching and checking the state uh and checks that nobody tried to steal your funds and you can think about okay if there is an apocalypse then maybe those fraud you know those challenges are not running however apparently those challenges challenges are running usually uh our co-founder Chris tried to check in the rollup that is no longer functioning it was uh you know um already disbanded uh on the first fuel uh rollup you know we figured out hey nobody cares about this rup it has been abandoned so maybe let's try to steal some funds at this um at this rup so we tried to steal some funds
and apparently we were checked we were challenged we lost the fund like you know one of the not so many cases when we actually happy that we lost the funds uh the last thing that we uh that you should check about your uh your rollup your mechanism is the exit window like this actually relates not to the fact not not to the question whether you can retrieve your funds after the apocalypse but rather if something bad happens with your uh mechan with your system with your L2 for example it's being upgraded in a way that you don't like for example government some government uh enforces um this system to censor you for any reason do you have an
do you still have some time to actually exit it and retrieve your funds before it gets updated and I don't have good um I got don't have good news for you uh for now I think that this will you know come with time but at this point most of those systems are upgradeable and if they are upgradeable actually it's hard to make any assumptions about this security if they can be upgraded in time so all of this is obviously quite you know complicated uh and you have to figure out all those different pieces so vitalic two years ago proposed that maybe let's simplify it let's maybe simplify it for the way to provide
people an easy way to assess whether this rollup is actually secure and mature uh and more importantly to push those rollups forward so that they actually Implement those things that are important for the security of our our funds instead of doing some you know fancy uh fancy features that are um good for marketing and we did Implement that we did implement the stages the we currently have uh two St three stages uh the DRS can be in um most rollups are unfortunately at stage zero uh the reason why we have it is to push those rollups to actually improve and it works actually last week arbitrum like these
are the announcements that we that we see rollups do care about this assessment rollups do want to improve with their security and they do care about this assessment uh this was this was arbitro announcing that that soon enough they will introduce permissionless uh permission fraud proofs and it was followed by the decaration from optimism uh they decare they declared that they will join the stage one rollups pretty soon and after that their goal will be to achieve the stage two and we are we are really looking forward to it because we want to distinguish good rollups from not so good ones uh even more than we do right
now and you know that's not it there's one more thing actually there's a lot of more things we have um value locked analysis for you we know not not not only show you how much value is locked in those different those different l2s but actually what is the composition of this tvl because this is super important if you've got a token on a c L2 it makes a huge difference whether this token has been natively minted on this L2 or was it um canonically bridged using the canonical mechanisms in the rollup to L2 or was it externally bridged through some
different external bridge to L2 these are three different animals those tokens uh and we not not only provide you the breakdown of the tvl for per category but we also show you where where is this number that we are looking at coming from we provide you data on activity this is actually a bit more complicated we are going to investigate it even further because this shows only TPS right now but you know with multic calls with account abstraction activity is a bit more complicated topic than just counting uh transactions uh we are recently introduced da assessment so we are providing you with information on
the um security RIS and composition of the those different da Solutions we provide you information on liveness we provide you information on finality why because this tells you whether your system um will allow you to function basically will will not try to censor you uh and if it tries will expose it uh last but not least recent like this is an addition that we introduced literally last week or even this week uh it's a cost breakdown we analyze already how much are those l22 paying for the security of ethereum to the ethereum and we provide you with this uh breakdown
and the question is why do we do it what is the reason why why do what do we do it for like just to call out uh those protocols actually no the reason why we are doing it is that we believe that this is not how you protect user funds stupid lawsuits and trying to put developers into jail for writing code is not the way to protect our funds this is not the way to go the way to go is actually to provide us with the information to keep um projects accountable to verify verify the their claims and to check whether they are actually providing with the security that they promise you when when they advertise the the offerings that's why
when we are analyzing those systems we don't just check the documentation we don't just you know uh look into what they claim we check into the code and everywhere where it is possible we try to back our statements with the uh links to the code where this particular um thing is implemented that you can check yourselves whether whether we are not wrong or whether we are not mistaken so this is our like motto trust the code and not the claim and you know with that in mind I can like we are doing this to be able to tell you Anon that yes your funds are secure uh however in order for you to be certain you have to go to our
web page and check actually if the system that you are using actually implemented all the mechanisms that they it was supposed to implement in order to be secured by ethereum and that's what we do for you that's something like we are public goods company we do it you know as a as a public offering so we are asking also you for support you can either donate to us and you will like below this on our website you will see the sources of of our funding uh we we we try to be fully transparent uh but if you don't want to do this we are asking for one thing like just go to our Twitter interact with us you know follow us uh like our post
comment on our post provide us feedback and spread the word you know tell your friends that they should be actually checking this data because it is important and last but but not least if you think that what we are doing is okay and that we are doing a good job in the e EOS system you can also delegate to us because we are present in the governance of those different elt and we try to push them also through governance to actually you know deliver on the promises that they made to you uh regarding the the security mechanism that they were supposed to implement and that's it for me thank you very much do we have time for some questions I think that we should
have does any anyone have any questions yep um so it seems like you guys are tracking all of the main um layer tws right in different types um can you speak to the different types that you guys have observed throughout the various um l2s that you guys are tracking obviously we know of the main optimistic rollups ZK rollups um are there any others to speak about um and then are you guys even like gauging the uh the differences between those different types of rollups in sense of like security um uh trust trustworthiness right of ZK of zero
knowledge technology in general right it's kind of like termed as moon man yes so like you know we are digging deep into that and there are it's a very good question because there are very a lot of different animals in this field like you mentioned optimistic and ZK rollups but that's not enough you know um there are different constructions like validium where the data is sitting somewhere else uh like optimums which is optimistic rollups with the data somewhere else uh there are different differences between those two as well and actually right now you know this space getting getting more and more complicated something that I had uh in Denver which I likeed very much is like somebody was Shilling uh orbit chain you know arbitr morbit chain that settles on Celestia sorry that
settles on Bas using Celestia a DA and uh espresso sh sequencing and you know with such constructions how do you actually assess the security of the system uh this is something that we are struggling with uh we already you know that it is already a challenge with l3s how do you actually assess the security free we are actually working right now on providing more information and more um like better assessment of those L3 um and for each for each of those con instructions that uh I mentioned you can actually see on our website let me uh find some screens screenshot uh so if you can see here yes
we do compare those uh those uh different solutions and in the values of this assessment you will see the details of the implementation for this mechanism in the particular system so for example for validium um in the Des data availability you will probably see uh that it's an uh that's DAC uh data availability committee usually uh you will see for systems that are using Celestia or um uh like I da like mantle you will see that the uh data availability is external uh for the systems that provide you with um es escape hatch uh you will see uh on the um sequencer fail uh column that they
they do provide Escape H and for example for dyx we actually implemented the interface um to to to do the you know to to escape your funds uh if you go to Dy dx. el2 be.com and it's especially important because dydx actually resigned from their ethereum based um system they moved to Cosmos so but if you have funds loged in your contracts on etherum you will be able to retrieve your funds using escape hatch from our uh interface so you can go there um and check it there so I don't know if this aners question fully but yes we try to analyze as deep as possible those systems and to
provide the overview to let you like you assess the risk yourself so you make a conscious you know informed decision when using this system or another any questions any follow-up question I think all good thank you so much okay thank you very much