DAPPCON 2024: Gnosis Stable Fees With Gas Futures - DappLion
DappCon 2024 · May 2024 · 11:29
DappLion speaks at DappCon 2024 under the title Gnosis Stable Fees With Gas Futures. Gas fees are the transaction costs paid to use a blockchain network such as Gnosis Chain, and a futures-style mechanism is a way of setting a price in advance rather than at the time of the transaction. The talk's title frames gas-fee stability on Gnosis Chain around that kind of mechanism.
Topics
From DappCon 2024, hosted by Dappcon Berlin.
Transcript
hello everyone happy to be here I'm just celebrating the bull market I hope you are as excited as I am so thank you for coming I'm d lion I'm the for coordinator of nosis and today I'm going to talk about a proposal to bring gas features to nosis so why do we want this as you know both nosis and ethereum has the same instant auction mechanism and that's very spiky like most of the time osis is a lucky chain that doesn't have very expensive fees and that's good it attracts the type of applications that that we want this stable value more Community Driven but it kind of discourages them at the same time because they rely on this cheap Baseline that can disappear anytime that I don't
know something crazy like this defi or nft crazy thing happens and everything spikes so this spiking mechanism is okay but it affects everyone the same way so let's explore what would happen if we would allocate block space in a different way since that people perceive these spikes not equally so our goal for this proposal is we want to expose a way that you can purchase the right into having a price guarantee and as a derivative an inclusion guarantee one gets the other because eventually everyone will be priced out except of you so today nosis implements EIP 1559 where it's really a good mechanism to expose a way for users to not have to
like think to much about price but this only works in the short term in the in the say minutes to hours fees can still Spike very bad and this does not satisfy our requirements from before so the main difference between ethereum and nosis and the whole thing that allows this proposal is that in ethereum we burn in nosis we collect we collect into the is collector bolt governance by the Dao I think today we're not doing anything for it one of the potential applications is to buy back gno so it's kind of an indirect burn but it can be used for this purpose so let me first give a shout out to Martin this is his idea block rent
I'm just extending it and definitely gas token was great inspiration these guys allow the chain to store gas in a way that's perpetual and allows some really interesting market dynamics but in a way it's not optimal to what we want to do because this token is permanent so you have infinite competition for the same block space you are just moving the current like spiky Market to another dimension of these token holders not ideal and padot is doing this thing called bulk core time which inherits a lot of these ideas that I'm going to talk today where you can give a specific entity person holder of some rights the option to have access to some sort of block space in a regular way um so let's
look how it's going to work so we could build this gas option without changing the protocol at all just with some trusted intermediary or clear house so this is how it would work in the real world uh like we could build this today just on top of the corrent infrastructure but we will need someone so let's work through the thing so you have the Speculator that buy is a contract with that with the app the app promises to buy gas at the spot price and then these two guys have to figure it out to settle the price difference so this last step is what requires trust collateral or something to make sure that this transaction happens and they don't bail on each other but we can do better if we change the protocol we can
have a trustless solution because the chain is aware that this option exists so what we want is you buy the option to write like the right to purchase end gas at some strike price during some window for a duration of Time how it's going to work you go to this system contract let's call it the collector contract and you buy this nft that gives gives you um rights to this future contract every window you can send a payload to this contract and this contract will forward this payload to like whoever you ask it meter and then refund you the base fee so effectively from your point of view you have an a base fee exemption you
only have to pay for the priority fee so this schol this works but this kind of defeats EIP 1599 in a way the key thing you have to understand about 1559 is it's more of like psychology to the users they have to trust that this mechanism works and that no one is gaming miners are not gaming and definitely there is no like collusion between Miners and users to to defeat this mechanism so to prevent that we're going to limit how much of the total block space is redeemable to here I'm picking 40% but I really want to run some simulations and definitely study this this problem to make sure that we are not unsafe I mean otherwise uh it's
not not good we can we can we can break this in Varian so first version b z roll out uh we'll start with uh commit reveal auction this is a classic mechanism it's very well understood it's used in ens and a bunch of other places like the just to recap the primary sell of these sorts of mechanisms is what makes them really annoying if you have heard about execution tickets it kind of implements something like this but how to you design as as like stable and solid primary aution is really really challenging so something simple like commu re aution should be fine we can explore a DCH auction but it's very inable and I think it can get really
spicy if you have like the mechanism that allocates who has right to exemp the base fee on a mechanism that can be gained by the same people that that get these like and features so let's just not do that for now uh we can start with some conservative parameters so just allocate 1% of block space per future and maybe 10 so 10% of total block space can be competing for for for these sections strikes price zero so we don't have to figure that out but uh I think easily we could Chone these bar contracts or make the chain Ware of historical base fee prices so that the strike price is the historical Bas fee it's a somewhat decent is approximation of the strike price and will make the
system much much more economically efficient because otherwise when the app buys this future gas contract it has to front run all the capital of the gas that will have to expend through the duration of the contract which can be a non- negligible amount and then some rolling window here the parameters of window and duration depends on how much total gas is competing for this just so you have an intuition in the case of gas token this rolling window is infinity and duration is also Infinity so everyone that holds the token is competing for the same block space and that can leave to this unwanted congestion so again we want to do more research on these parameters but this is like a good good first uh values
that that we can uh deploy for v z so just to put a diagram see how this is going to work so you are the app there is this running monthly auction for these nfts you hopefully win it then you get this mint nft which because it's nft you can have like this secondary order market so the app may not be the primary purchaser could it could be some Speculator someone that understands how demand would evolve in the chain in the future and acquire this nft and then you as an app that come like later after this primary sell you can acquire this these nfts maybe not only for like month zero but you can get one like one for month zero months one months two months three and you have basically like a guarantee of a year worth of gas in a predictable uh price then in every
window you send this transaction to this contract that will meter how much gas you spend and then reform you the base fee I want to note that we try to use the ERC 4337 standard for this but unfortunately it doesn't work the current entry point contract M meters gas properly so that would be fine but it reforms both the base fee and the priority fee and that would break the the whole architecture but we can we'll figure out we we'll do something in the future we'll see on the road map so this version with this simplish contract it's good for all these use cases basically anything that does not depend on a message. center could be used so say a
safe that's ready to be executed could use this like the safe team could purchase these gas fatures and have some assurances that they will just not bankrupt themselves the chain goes crazy cop execution gelato Network any any relay Network would great benefit from this could use it and also like Po meting like it's the prime example they have this constant load through all the conferences to to mean these nfts so that's B1 b0 what we want to look next uh we can definitely change the the block production strategy and a tester strategy s such that we guarantee inclusion at a consensus level so owning this nftd would not only give you this like price discount but it will also guarantee that your transactions are
included through some window or within some conditions which is not the case now second point I think we can do this on bz is having a higher strike price so that the system is like as as more as economic efficient as possible you not asking apps to front run as as that much Capital uh another feature that we inherited this from the pka bull core time if these nfts are splitable that's super interesting because you can have a much much more Rich secondary Market where you can just buy this 1% and then split it in whatever fraction you want and then resell it will allow much more expressivity of what the market wants how much block space at what price and
and at what range it will be like really interesting um I'm also I think what will be really EXA like excited about this this would become a prediction market for how speculators see and value uh what's going to be the usage of nosis chain one more month two month 3 months four months into the future and last Point as I said like the current version is not compatible with E4 like 437 entry point contract but it will be possible to have a slightly modified version also audited with a different address that people trust but we can use the same user operation standards to package these uh refundable uh
transactions and I want to point out I think this is a really really interesting opportunity for nosis chain because no other chain has this property of eap1 559 collection in a way that's like credible and trustless and this will allow nosis chain to have like an edge for this type of users or applications that wish to have some like pricing stability looking into the future thank you so much [Applause]