Multi-Dimensional Price Oracles | Jakub Wojciechowski - RedStone Oracles

[Music] hi everyone uh my name is Marson I'm co-founder of redstone oracles and uh most likely you've seen on the agenda my other co-founder Jacob who unfortunately yesterday couldn't board the plane so I will be running the presentation and also there is a slight change in the topic so it's not going to be super technical because I'm not like a coder myself uh but it will be mainly about roll up as a service and oracles and I would try to bring down a bit more the concept of like how oracles differentiate from each other and why would you have another Oracle probably than chain link that the majority of the market is very much familiar with so throughout the presentation I will try to ask some questions and my idea was to give out some Redstone merch for people to uh uh who who would answer that but I think it will be probably a bit difficult with you know the space over here so just after the uh presentation if you would like to get some Redstone merch like catch me up and I'm happy to distribute some socks and other goodies so uh no further ad du oh I'm also co-founder of eoro which is a conference in hackaton in warso Poland which I recommend joining uh this year a bit about Redstone before we kick it off so uh we started late 2020 and uh came with the final version of the product in early 2023 so January 2023 was when the First full production application started using Redstone it was Delta Prime and since the Inception we managed to close $8 million in funding uh the seed round was led by lmis C and had some good names like combus Ventures blockchain Capital Maven 11 also joining and believing in our vision in uh disturbing the Oracle space we also got backing from uh one of the most prominent angels in the space including Stan kleo founder of AA and deep from polygon Alex gloski founder of ZK sync or Amin founder of avalanche one thing that you're going to notice many of these Founders have their own like ecosystems chains and they understand the necessity of diversity with oracles especially not having like just one provider to be dependent on we are we're pretty recognizable right now in the space with our reports we've published five so far so every quarter we do a deep dive into a sector that relies heavily on oracles and try to bring down the main players the Dynamics of the sector and what kind of a role Oracle oracles play in there so we've done stable coins uh real world assets uh liquid staking tokens which was by far the most uh read one uh it had over 500,000 views so far liquid raking tokens and H crypto yield and earn products so if you want to check them out you can go to our website or Twitter uh they're fairly available and it's I would say a really comprehensive overview of the particular uh part of the defi game that we described in there and we are a team of 15 cool people actually I mean like we care a lot about the culture fit and also being a professional and also going when you join Redstone so right now we are uh 15 always with uh Redstone merch like vest socks and other stuff um yeah so this is a bit about ass before I jump into like the different types of oracles maybe first why you need an Oracle in general in defi in blockchain crypto right so oracles deliver data to DS for smart contracts execution it should be as Broad and as simple as that but actually if you are Prof efficient within the space you will quickly realize that currently it's mainly Market data so volume interest rate market cap and so on because right now crypto is Major mainly around Defi and these are the statistics data that defi protocols need to execute properly and then if you're within the Oracle space you're going to quickly realize that you have to narrow it down even more because actually right now as it is of today February 202 for over 95% of the data delivered by oracles are price feeds of assets and mainly crypto not even like assets from Real World I mean more and more from Real World but majority of that is what's the price of Bitcoin what's the price of ethereum what's the price of usdt and so and and so forth for for def protocols so you are going to quickly realize that oracles is like a super broad concept like okay bringing data on chain cool for execution but then the implementation is so hard and also so expensive to create a robust decentralized and resilient Oracle that currently the product Market fit is mainly about price information about assets mainly crypto and I wanted to ask a question what is the most utilized price feed right now in the whole defi but can someone try to answer if you know all right I think it's pretty difficult to hear but it's if USD so as it is of today 90% % of the queries when it comes to price feeds is on eusd and the major reason for that is because ethereum is the biggest chain right and then many protocols even if they require another price feed they denominate all of that stuff into Eve so in many places in their system the denomination between if and USD is crucial right and right now we see the biggest increase actually in derivatives of e so liquid staking tokens and liquid reaking tokens because they are verying guil and they are used many times across various lending uh CDP options and so on protocols so what are the use cases for Arles when I talk about defi um oh yeah so actually oracles will be at one point used widely in web free ecosystem that's probably given but until now it's been mainly focused on defi so price feits at defi these are oracles as of 2024 at least early you use them in landing markets such as AA or venu on many chains CDP stable coins the most recognized one are maker Dow liquid but are also like a rising stars like Prisma leverage protocols so gearbox Delta Prime Kean and more derivative protocols commonly also referred to as per protocols or perss so GMX uh especially right now if they're V2 or synthetic perss or Vertex or other Perpetual platforms are using oracles for pricing the assets that they're allowing to trade yield and earn protocols so right now would say the major players are SAR enzyme uh defi saver instad app and so on so these are the protocols that also uh rely heavily on the price Fe from oracles and algorithmic stable coins like uh Mento or fra can you come up with two defi protocol types that doesn't need Oracle because I'm talking about many that do need so if you turn the tables it turns out that dexus usually do not utilize oracles so like uni swap Sushi Swap and so on uh the main reason is because people trade on these dexes and actually they're it's the reverse so oracles many time rely on dexes as price fed sources so they catch what's the price of the trades and then aggregate it from many sources and deliver uh to the protocol um if you can call it if I protocol a bridge then they also do not rely on oracles and there's probably some others but if you extract like three or four categories then all of the rest of the defi do need oracles sometimes there are new architectures like with Anja uh that try to create Landing platforms or defy Oracle less because oracles if you can create a protocol without Oracle do so I mean like oracles are just another piece of the whole puzzle that can go down or can be manipulated or can be something wrong with it but majority of the times if you want to keep the app um fairly with a well with a good ux and also performant and competitive on the market you do need to rely on oracles so why do oracles matter for Ras so Ras stands for rollup is a service and it's um a service where a party just requires a rollup for their specific use case may it be defi gaming or something else and many times when you create a rollup you're like okay let's go to one of the providers like conduit outlayer gelato Caldera and you create your rollup and you like hey I have my own chain it's cool but it turns out you need also infrastructure on top of it and oracles usually are a very essential part of this uh infrastructure so if you want to have dii on your RAB you need an Oracle and you can probably just call chain link hey guys like can you please Deploy on my chain like I just have it deployed like in two weeks can you come over over there then you might get an answer of course like I will put you in the pipeline probably in couple of months or maybe even longer don't get me wrong like I appreciate chain link a lot actually but for them and creating super robust system that almost never got truly um like malun has never malfunctioned uh we wouldn't have defi as we know it today so we do need chaining and it's like a super important part of the whole ecosystem but there are uh not that many sectors in defi that are as monopolized as as oracles and it's like a critical part of the defi stack so we need more diversity in there so I would be asking like how many chains does chaining support and it's right now roughly 13 14 chains uh if you dis like that take out the Legacy chains because there are some it's more like even 12 and then you realize that in current War we have probably like 30 let's call them viable chains 50 maybe promising chains and then the whole sector of rollups the service so new chains that are popping out every other week or or month so uh you see that not the whole spectrum is covered by the main Oracle provider and actually the another question would be do you know any other D5 vertical that has been that much monopolized as oracles so some people claim uh dexes with Unis swap but actually I would say right now with so many chains and so many various flavors of creating a DEX and also like if you add per dexes it's not that monopolized anymore like there is innovation with wallets of course like metamask is the main one the main used one but the competition is like very Fierce over there and so my claim would be there's probably no other like D5 vertical as heavily dependent on one provider as oracles which in our stand should shouldn't be the case probably um so then the question why chain Ling is not on the all the chains like what's stopping them from being present everywhere and joining the RAB the service very quickly and efficiently so the key bottleneck when it comes to Oracles in general as as we know nowadays is that onchain data storage cost is very high and monitoring infrastructure is also like pretty expensive so the key consideration that each Oracle is taking into account is the update condition and up update of one up and the cost of one update so how many times I need to update something on chain and what's the price of each update you can multiply and then you can get a cost over for example a month so there are two main implementations visible on the market right now and chain link also has both of them we also have both of them and other oracles also tend to offer either or both of them which is push oracles versus pool oracles and for anyone who over here who really would like to be within D5 and understand the space I would say uh you should know this concept because it's going to be repetitive over and over for the next months so what's the difference between the push Oracle and the PO Oracle and I will try to break it down in a second so the main idea of the push Oracle is like the infrastructure that chaining brought to us 2019 2020 and popularized is when you aggregate data from various sources say it is decentralized Dex centralized uh exchanges uh price aggregators like Kiko coin market cap you aggregate that in a a layer for chain link it's done decentralized Oracle Network and then you shoot that data on chain into the storage with intervals and these intervals are dependent on a time interval or a specified price deviation what it means is for example when you have if USD price feed and you want to update it on chain you are going to update it for example every 24 hours so that there's like a heartbeat or the second condition is the deviation so it when it deviates from the previous update for at least for example half per or 1% so you're not you're never going to have real time data on chain because it's just too expensive so you're are accepting that there is always going to be a specific um like AC acceptable difference so in that case if it's for example 1% so so you can have a difference in Arbitrage of up to 0.99 of percentage point for if USD on chain and when you add to that the volume that such a feed is you know protecting then the Arbitrage is pretty big in pool oracles the main idea is to update data on demand so when there is a transaction of a user that requires a data fit from an oracle then you deliver that data on chain so only when it's truly necessary and the transaction can go through without the Oracle data only then you're delivering that and that's the main concept and idea of a pool Oracle so usually the user interaction is the update or trigger condition not like an interval or um hardbeat right and now why is it important if you create an oracle that is just dependent on the hardbeat then it's per fairly easy to calculate how many updates you're going to have for a month and then the only variable you have is the cost of one up update because we know that the gas on various chains like Devi like deviates all the time then if you add the um the deviation threshold then you cannot even predict how many updates you're going to have throughout the month so you have no idea what your cost of running an oracle is going to be is it going to be $200 is it going to be $500 or is it going to be like $30,000 with the gas spikes on ethereum sometimes it can be even like crazy number like in millions like believe me we did calculations and operating like chaing oracles on ethereum in the moment of a heavy gas spikes can cost even millions of dollars weekly that's crazy that's that's insane but this is like the cost that you have to bur so that the price feeds are updated on chain in pool Oracle you can also you you also cannot predict how many updates you are going to have because these are the users that interact with the DAP that are going to trigger the updates so if it's like for example night or okay there's never night over over the world but let's say it's night in the heavy traffic geographical regions or the market is just you know pretty flat and there is no much movement you're going to have just a few updates but if there's like a heavy uh movement market like I don't know uni token goes up like 30% within like hour or something else people usually trade they try to catch like like an opportunity and then you're are going to have many transactions so many updates but here is one uh very important caveat is who covers the gas because it doesn't have to be the Oracle that cover the gas in the push Oracle usually it's either the Oracle provider that covers the gas of each update and then either subsidize that and that's what chain link has been doing for a very long time like they were been subsidizing this update so that is like more like a public good or it can be a protocol either the one that consumes the data or the one that the price feed of a of an asset is updated for so for for example in let's say ideal word probably the gas coverage for updates of if USD should be covered by ethereum Foundation because ethereum is like the coin that they are kind of like keeping so they probably should be the sponsor but you're not going to charge the ethereum foundation they might give you a grant with new protocols right now we see that pattern that it's in their interest like in the interest of the coin issuer to have the Oracle running because that creates a market on defi for their price feed they can go to AA or other protocol compounds Prisma and so on and try to Lobby to other uh price like like an asset as a collateral in the protocol so it's in their interest to cover that whereas in the on demand and pool model that's super interesting because usually you create a system where the user is actually covering the gas cost right now the majority of the cases the user is paying additional gas on his or her transaction in order to update the price on chain so the logic is more like hey I want to interact with this protocol so I have to accept the gas that it costs to interact in general plus the gas cost of the Oracle update because without an oracle this protocol wouldn't wouldn't work and then you you verify whether the users are actually willing to pay the gas and the the whole game is about how much gas it adds if your regular transaction costs $15 and you're going to add a gas cost of like 20 cents probably majority of the people are going to accept that but if your the the cost of your transaction is on I I don't know BNB chain or arbitrum or something cheap 20 cents and then you add additional 20 cents for the gas upd for the Oracle update which is like 100% increase in the gas cost you have to pay then it might be a problem is it like justifiable or not so many oracles including gas try to optimize the gas cost of adding these price feeds to transactions of users there's also like a new thing that Oracle started like experimenting with and of 2023 and beginning of 24 is to create a kind of like an accounting system of how many updates have been done by users of that protocol on chain gather it throughout the month and then the protocol pays a re embarrassment for example to the Oracle for all the updates that have been triggered by the user so there's always the question like who is going to pay this gas because this is the most unpredictable cost and it can deviate truly a lot believe me it's like between $50 and $50 million like in some time frame so it's it's a very pain in the S when it comes to going cross chain push Oracle is non-scalable the reason for that is because you have a service that is delivering data on chain imagine it's if USD and it delivers on chain with specific like conditions right but then if you have another chain and you want to have the same data on the other chain you cannot just you could potentially like query the other chain but there would be like a latency of quering the other chain and actually is going to be expensive as well so for the same price feed if USD you have to shoot that to one chain to the other and every other that you want to deploy to so here you can quickly realize that it's not scalable because the same price feed has to be updated on many chains and there is like bigger and bigger complexity because the chains are not identical many times you hear it's evm it's the same actually it's not because you have different RPC provider you have different like caveats in the code like how the guys are actually made it evm sometimes you have problems with signatures and y yada yada so people try to make something evm but usually there are some trade-offs if it's like like not true like native evm and then you have non evm chains which makes it even more complex because with each chain you have some other like configurations within the ecosystem to make the updates on the given ecosystem with the pool model many times you try to create a system that because they use like for the Oracle it doesn't matter it shouldn't matter like to which chain the user is sending the transaction because you try to trigger the or Oracle update with the user transaction so with redstone what we created we created a flow that works out of the box on all evm chains the reason for that is we found uh that on all EVMS you have call data in each user transaction and you can pass some message in call data I'm going to talk about that in a second and that's what we we use some other oracles use for example Bridges or other passing services to deliver data onto many various blockchains so by principle it's just more scalable with the full and On Demand model and when it comes to number of feeds that's actually pretty connected with the previous point because okay you can have like just one price feet if USD but if you want to create truly defy you probably would like to have the number of feeds like I don't know at least 10 or maybe 20 or maybe 100 and then if you think about push oracles where we have to push oracles on chain each single price feed is going to have its own heartbeat and deviation threshold because it's not like if if different like deviated like half percent doesn't mean that Bitcoin deviated half per. so you have to create one monitoring for these deviations for each single asset and update conditions for each single asset so then you get a matrix of many to many with many chains like how you're going to update them and then you also learn that the chains don't necessarily want the same price feeds so everyone wants if USD that's given but when it comes for example for AA token or uni token it doesn't mean that each every single chain is going to need the same price feed so you have even more complexity in there when it comes to pull Oracle usually you keep these price feeds offchain in any sort of a constract or another blockchain from which you query so it's way way easier to make it uh scalable in the sense of the number of prices you support so sometimes you can even do like 100 or 1,000 price feeds without that much additional fixed cost when it comes to the destination chain so how does the standard push Oracle model flow looks like you have data source ources as I mentioned decentralized exchanges centralized exchanges or price aggregators you have node operators or data providers people call them differently like T-Mobile or flow Traders with a market maker or Kao that aggregate that data and then you have a smart contract on the target chain towards which the data is pushed with these conditions I mentioned before so say it's ethereum arbitrum avalan Solana and so on and so forth and then when you have a Dap that is deployed on that chain the smart contract of the DB can query the smart contract of the Oracle and get the freshest let's say price feed so this is very elegant and very simple and this created the defi as we mostly know right now but if you want to like scale it it probably is not going to make it affordable at large so it's not scalable across chains and number of assets how we perceive that is a bit like um early time email or like the like we we still use emails right like everyone uses that and they haven't changed that much since the Inception so like it's a very important construct but we also created many new um concept of messaging between people like telegram WhatsApp and so on so forth so our view is push model is going to stay over there and it's going to play super important role but we also have to innovate to make this um like boundaries not relevant so for spefic spefic uh types of applications because in our view updates of 24 hours and half% deviation is not going to bring us to the new Financial system as we all want to build here in defi so the pool and On Demand Oracle model looks a bit differently on the left for you actually looks the same so we have data sources and data provider nodes so that part doesn't change but what changes is that these data providers after aggregation or Price Discovery they do not shoot data on chain directly but there is like a sort of a middle layer that is way way cheaper than than um pushing data on chain because it doesn't usually contain gas or if it it does contain gas the gas is marginal so they sign these data packages and send it to a middle layer in Redstone case we call it data distribution layer or for Simplicity it's like an offchain cache so a set of gossiping notes that keep signed data packages for short period of time so that it's available for anyone to pick and is redundant like you cannot DS that for piff for example another Oracle provider it's piffet which is a Solana Fork which they govern they have 100% of supply of the tokens so they can fairly easily uh look like what's the gas price and they pay it actually in the native token so it doesn't matter for them and other oracles they have have their own like middle layer as long as they provide the pool model and then and when there is a user transactions to the trigger then deliver they the deliver the data feed onto the target layer one or Layer Two so for example for PNET they rely on a bridge which is called Wormhole one of the biggest players on the market uh it can be actually picked up by a liquidator so not only oracles deliver price on chain like if a liquidator is uh allowed to deliver a price feed in order to close a position on a or other Market they can also like deliver that so a liquidator can be like a party that brings data from this middle layer to the onchain environment or in our case in Redstone we were like okay users send transaction on chain anyway so probably the most efficient gas efficient and also elegant way to do that whole flow is to piggyback user transaction that goes anyway so you don't have to trigger anything else and we found a way to do so in this called Data if you want to check out how Redstone flow actually works on on like I don't know 50 chains like evm chains you can just go to showroom the Redstone Finance you can quickly connect to one of the chains and then see how the sample data is actually delivered in this call data field so why I'm repeating this call data call data call data because many people don't know about this construct and it's super important in my view so call data is read only by addressable space for example in ethereum which is like a state machine you have state one and then state state a and then State B in between you have a transaction in the transaction you have of course like information from who to who and the value but you also have data this called Data where usually you pass Arguments for the function so here is for example uh Charlie from the second field and from the Third Field and two from the second field so these are like arguments but you can add any arbitrary data as long as let's say it's gas efficient so our founder who used to be smart contract auditor with open Zeppelin and has been in Theon space since 2015 created a very optimized evm connector library that we utilize in order to attach designed data packages from this middle layer to the transaction of a user so it that goes to the destination chain this is actually from ethereum white paper all right so when you're creating your new chain with roll up as a service or not necessarily you're just launching a new L2 we've seen many recently like blast mode a fra sty right now there are a lot of l2s on top of Bitcoin so all of these if they want to play with defi they need oracles so they usually have to take into consideration like which Oracle they would like to have on their uh on their ecosystem and the consideration they have again is like the push you have the intervalent deviation pull it's user interaction that's actually the same uh table as before so from the lens of a L2 like an ecosystem or roll up as a service provider you have to take into consideration the waiting time for the deployment so how long do I have to wait for the Oracle the cost of deploying of the Oracle so what actually I'm going to pay and was the maintenance and types of applications that I want to have on my ecosystem because some applications work only with the push model some applications work only with the pool model and some applications actually can work with either and in my opinion like if you can create right now from the scratch a protocol that is going to be able to F fetch data from pool or push Oracle depending for example which one is the freshest or the most accurate this is in my opinion at least the near future synthetics is working on something like that there's a protocol that is launching so called cven that created such a system so it's like a new way of quering the data from oracles that is let's say the most fresh and the most robust SO waiting time with pull Oracle is always way longer than pool oracles because you have to create this whole setup for shooting thata on chain the cost is also higher so that's the reason you you have to innovate and right now I have to be frank majority of applications are just used to push oracles and even though in their flow they could include pull oracles they prefer not to because they audited the code or did something else and they just wait until the push Oracle is available so this is this is the state as it is today but in my opinion as we go forward with all the limitations I meant before the developers will be innovating to incorporate more and more pull or on demand oracles into their system actually there is one more consideration when you're creating a new chain hey probably you're going to have your own token right so if you're creating modes there's probably going to be mode token blast token we see arbitrum has ARP Optimus has op so there is probably like a token with the chain that you're creating and for you it's going to be essential to have a price Feit for that particular token on your network so with push oracles It's tricky because if you are going to configure let's say AR price feed to USD most likely other chains are going to use that as well because the market capup is really high but if you're just boost trapping the network there is no much sense for other chains to need that price feed so it's going to be super custom just for your particular ecosystem and many times in push oracles the cost is the cost is just un unjust you cannot justify that so uh couple of case studies when it comes to RAB service providers uh there is a RAB service provider called Caldera that has many types of stacks like op stack polygon stack Arbitron stack and uh Manta Pacific is one of the chains that they helped bootstrap they started and right now is running on op stack with Celestia as da layer but they will be switching over to polygon ZK evm and it's a pretty good success story right now Manta itself has I don't know probably like $800 million in TBL and uh the ecosystem is still uh growing um you can also give life to the old chains that everyone thought we're not going to make it so Lisk H if there are some crypto G's you probably remember like Lisk from uh back in the days in other times like very smart team I I got to know their team personally but in the recent months they probably haven't been so much active but right now they partner up with gelato which is another rooll up as a service provider and on top of op stack they're going to be transforming from a layer one into an L L2 on top of RAB a service provider and actually they're going to need oracles uh cell is transferring also from layer one to Layer Two without a provider at least for now as not as long as I know but the Forum of cello is super interesting if you've never been on a forum of cello I highly recommend you to check it out because right now what what's happening is cell announced it's going to transfer into L2 but they haven't done any like final confirmation like which l they going to like which stack they are going to choose so right now all the stack l2s like arbitr stack op stack ZK stack ZK evm stack and so on are creating a post on their on their Forum on cello forum and trying to convince them to use that particular technology so if you are ever wondering hey what's the difference actually between arbitrum Stag and op stack and so on I highly recommend going over there because these guys are um flexing their muscles just in order to get cell because it's a pretty big ecosystem already uh Kanto is another type of a blockchain uh that has been or still is L1 and want to wants to transfer into L2 so you see like there's a pattern of older l1s that realize that probably it's uh just re more economically viable to go into uh becoming an L2 uh maybe one day we are going to see EOS I don't know whether any of you remember the EOS Ico and knows this ecosystem but they the guys still have a lot of cash so maybe we are going to see like a Revival of that ecosystem in the form of a roll up as a service or another player parallel that also is going to launch or I think maybe already launched with arbitr stack uh utilizing conduit as far as I remember as a provider so I'm going to go into wrap up and one thing that I still want to leave with you is like a simplified framework or maybe couple of questions you should ask yourself when you are comparing oracles and you're trying to pick the one so number one is customizability can I decide on which data provider to trust so for example with chain link you are accepting the setup as it is you're just chaining is the one to decide like all the customizability aspects but with for example Redstone I believe piff and other providers you have more uh let's say gears to tweak in order to make it uh work for your particular chain or dab so you can for example say hey I don't like this data provider so can you please disregard it and then you just in Redstone case disregard the signature of this data provider and then you accept it from all the others so you have like the power to uh do the choice centralization so can the Oracle switch off the infrastructure and here it's like in my opinion there's no perfectly decent Oracle and in anytime soon we are not going to see one because it's super hard to create something that is fully decentralized when it comes to such an infrastruct architecture when it comes to chain link uh to the best of my knowledge there's like a multis I think like five out of 10 or something like this that is governing all the contracts H in Redstone we are also like slowly progressing in the decentralization uh but it's in my view especially in oracles there is no like centralized and decentralized there's a spectrum and you have to like look like at what part of the spectrum a specific provider is and if there's like a path to keep decentralizing the system redundancy so what has to fail for the or Oracle to go down so what are the critical factors for the Oracle to keep operating properly and uh evaluate what is the probability of uh this particular part uh to go down so for example if you rely on a bridge in a pool model if that bridge goes down actually your Oracle is not working and unless you have a backup system maybe there is maybe there is not so it's very important to check that out uh layer one layer twos and roll up a service compatibility so is it designed to be cross Chain by Design number and types of ass of data feeds available and how easy it is to add a new one so we see right now also like some oracles being super performant and optimized for specific price feeds so for example they offer only free Price feeds but let's say with very high speed or high performance but they are limited to add new ones uh on the other side you might have less performance price feeds but being able to add um like new new tokens as they as they go onchain versus offchain price sources oh so this is super interesting especially important in the beginning of 2024 so can the Oracle for example provide a price fit for a token that is mainly traded on chain on dexes because there is a suit of tokens right now that are not traded on centralized exchanges are traded on dexes and actually can make as a perfect collateral for many of lending protocols because lending protocol if the health factor of a position goes beyond a sergeant threshold usually tries to sell the asset on decentralized exchanges so the liquidity of these exchanges actually matter the most for them so liquid staking tokens is an example so R st e r if from rocket pool uh as swe from swell Os from stake wise um ifx from Stater and many many more and in the beginning of 2024 we've seen the biggest Boom for liquid reaking tokens so the ones that when you lock your if you get a liquid reaking token and the if goes to h protect igen layer and goes for reaking so the biggest players are eeri Renzo kelp Dow puffer Finance uh so these tokens work as perfect collateral as this for today on the landing markets so if an oracle pulls data only from centralized exchanges which is easier in in in in general everything that centralized and offchain is easier then this Oracle probably is not going to provide the price of an asset that's traded only on chain and the inherit nature of lsts lrts or also some other yield bearing assets is that they're traded on dexes Unis swap curve balancer you name it another one that is upcoming is a te as usde so actually the most liquidity especially in the beginning is going to be on dexas as well most likely so if you want to have early this you have to verify on your Oracle I haven't touched upon that one but it's also like pretty crucial was the price aggregation methodology so is it robust against for example wash trading so wash trading is placing like uh two traits for the same asset that are opposite to each other so that many times you try to omit you you you try to create a loss so that you can omit taxes or some other aspects and if you have a pool on dexes with low liquidity and someone creates wash trade over there and uh the volume spikes you can easily move that price and actually this is something that we have seen we have witnessed in December last year with chain link price feed Oracle for R stake if on arbitrum so in a balancer pool with low liquidity there was someone was trading 2 million ion doll of R Stak if and that moved the price because chaining is relying on vup so volume weighted aggregate price feed and for onchain let's say price Discovery it's not necessarily the best it's better to track liquidity because it's harder to skew and more expensive to move across uh many places or is it price aggregation is it median or maybe is it something else so it's also important to keep in mind if there have been any historical mispricing and manipulations for an oracle so has it failed in the past and how to understand whether they improved or not are there any guarantees so if the protocol gets any damage because of an oracle am I going to get get any damage payout because there are many models like through Redstone we also work on a system that pays out the damage to the parties that are for example um yeah got a loss because of the malfunction of an oracle govern tokenomics and audits that's the same for all the D5 projects so fundamentals and if there are any circuit breakers so what happens at Black Swan events for example what happens if FTX goes down or what happens if uni token goes up like 50% within minutes or hours so to see to understand also what happens in the critical moments because oracles have one job like be robust and be up so this is like the most important and to understand like how they behave when truly they are pushed to the limits because they cannot like fail in that these these moments I wanted to create the Oracle comparison table but I haven't managed to like the time didn't allow me but soon on Redstone uh profile on Twitter like Redstone un defi I'm going to finish that and publish over there so that users have better understanding of like how chain link Redstone piff apa3 and other oracles actually compare to each other and what are the major aspect that they differ differentiate from each other because also for you to understand our approach is that uh me actually all of the names they mention right now are doing fantastic job in their own fields and in my opinion it's super important to uh keep appreciating like for example chaining for delivering what they've done so far PFF also what they've done and uh also shout out if you see an oracle that is a bit more shady like making airdrop promises or I know advertising cars I'm not going to show names right now but you can probably check in Google like Oracle that promotes getting a car or something like that when the code is not speaking for the Oracle rather the marketing campaign then you have to know that something is cooking in the back and at the very end I would like to invite each and every of you to e foro conference in hackaton that is going to happen oh this year it's going to be beginning of September from the 2nd to the 9th of uh September at University and Wars University of Technology invite all of you very much there and thanks a lot for spending with me these 45 minutes I hope you enjoyed it and learned something and I don't know if there is like an option to get questions over here or not I don't think so if anyone wants to ask a question raise your hand okay there is a question yeah but I don't know is there another mic okay all right so it's pretty comt uh not comfortable over here to ask questions so thank you a lot and if anyone would like to ask anything I'm going to be here next to the stage thank you a lot