Last week, Loopring’s Chief Architect, Brecht Devos, sat down to chat with Chainlink’s Johann Eid and Rory Piant in a live video chat. They spoke about ZK Rollups, DEX scaling, and the use of Chainlink oracles to secure a component of LRC staking. The video can be found here:
In addition to the chat between teams, the community asked questions via the Pigeonhole app, but not all had time to be answered in the video. To leave no one wondering, we responded to most of the questions after the fact. If a question from the app link above is not answered below, it was either answered in the video, or effectively answered in the course of responding to others. We initially recorded our responses on reddit.
Q: Why do you favor an Orderbook model over a Person-To-Contract model as seen in the Synthetix ecosystem? What are the benefits to an Orderbook model?
A: Hi bad ger. It’s a good question. I don’t think we view it as so much favouring per se — they both have merit — just different markets, or market structures to be exact. But I guess at the end of the day, building one way means that is your favoured target :). As you know, Orderbook exchanges are how the vast majority of real-world assets, (or at least securities / other financial assets), are traded. They are effective at coordinating attention and liquidity around assets, and most people are familiar with them — from trading equities, or indeed, cryptoassets. The best answer I’d give is really expressed in a post we wrote a month ago that contrasts Orderbook exchanges vs the simple swap, on-chain types (Kyber/Uniswap). What we call ‘real trading vs simple swaps’. (I suppose Synthetix is in the ~same bucket as Uniswap re: P2C). But broadly, we are fans of the P2C model — which is uniquely enabled by blockchains! (and specifically DeFi on Ethereum). However, we think Orderbook trading isn’t going anywhere, and is how the bulk of financial assets will trade, so we are looking to be the most performant infrastructure for that market structure.
Lin K. Marine
Q: Are you gonna use chainlink oracles to facilitate external payments via SWIFT or leverage automatic trading routines driven by data in google cloud?
A: In terms of SWIFT, this is something we have not given too much thought, but it is certainly interesting and important. We recognize the need to bridge the external, legacy systems (fiat on/off ramps) and indeed, will be pushing into the product sphere (as opposed to pure protocol) in 2020. We have built a smart contract wallet firmly focused at onboarding Chinese users into the DeFi ecosystem, and most prominently, Loopring v3 DEXes. For automatic trading routines — we actually kind of referenced this in the video, and in our integration article with Chainlink — we most realistically will use their oracles to provide external signals for advanced order types (stop loss, etc), and more complex automated routines.
Q: Can you explain the usage of Chainlink nodes in conjunction with the loopring ZK rollups? Are they needed for each batch of transactions?
A: Hi Dr. Assuming you may have read the Loopring — Chainlink integration post, (or if not please do :), you’ll see a Chainlink node is not “needed” for our ZK Rollup to work. All that is needed for that is Ethereum, and Zero-Knowledge Proof cryptography. That is all the Loopring protocol builds on, and how we maintain 100% Ethereum-level security. Chainlink factors in by helping out with a sort of ‘interim’ level of economic security, while the zkSNARK proof is being generated (because it is not instant). LRC is staked to ensure good behaviour by the Relayer *while* the batch of trades are not cryptographically proven yet. While not ‘needed’ for our ZK rollup, it’s still a big part of the design.
Q: If we stake LRC, how long do we have to wait until we see any rewards?
A: Hi Juan. LRC staking has minimum period of 90 days. After which, stakers can withdraw, or roll it over, with the rewards they earned from the past 90 days being added to their stake. So technically the answer is 90 days. But you may be asking when can you “really” expect any rewards :). Of course, it is non-deterministic and primarily dependent on volume running through the Loopring protocol. However, it is not solely dependent on volume — LRC will also be rewarded to stakers from the economic bond which is staked by DEX owners (the type of staking that Chainlink’s price feed is used for). When a DEX spins up, they must make a minimum stake (250k currently), and this goes to LRC stakers. So, by virtue of WeDEX, and the coming Hebao DEX… that is an effective 2% yield by March based on the current amount staked by users (23 million LRC). Please see the very bottom of this post for more info on that.
Q: What kind of performance tests have been performed, and were these within testnet or mainnet? Are any objective reports available which thoroughly document methodologies employed and the results?
A: This one was answered by Brecht on Twitter :). You can see how we came up with our throughput numbers in our protocol design document. We also have actual numbers currently achieved in a closed beta on mainnet: deep dive of the data.
Q: Is something like high frequency trading possible on a DEX? Are there analogous exploits that can be carried out on a DEX running on Ethereum? (That is, exploits that are unique to such a platform.)
A: Hi Sasha. Good question. Up to this point, no. Part of Loopring’s whole push into scalability with v3 is that real HFT was impossible on Ethereum DEXes before, because they were so slow/expensive, as you likely know. Now, I’d say it is becoming feasible. I may be misunderstanding your next phrase, but one small point — I wouldn’t call HFT an “exploit” outright. In fact, that is a perfectly legitimate (and popular) strategy for several traders, and especially market makers/liquidity providers. So v3 making it possible is a good thing :). We need the speedy traders to be able to trade on DEXes if we are to overthrow the centralized ones. I feel you may be referring to something associated to HFT which can indeed be nefarious — front running (am I correct to infer that?). Front running — or jumping in front of others’ trades after you’ve gathered some info — can be bad for markets, and dissuade participation. You’re right — that was/is a BIG problem on many (~all) on-chain DEXes such as Uniswap, or Radar Relay, because the intent of the trader is perfectly visible for anyone to glean and trade against! (by increasing their gas price to fill it quicker). You’re also right this was unique to DEXes. Loopring v3 does NOT face this problem, however, because orders are not placed/filled on chain, but via an off-chain Merkle tree, and then proven in a ZKP, which is then submitted on-chain. Thus, only the DEX (Relayer) sees the trades, has the ability to match them, prove them, etc… no unknown party can front run settlement at all. So in short, HFT possible, front-running is not :)
Q: Who are the other oracle players in the decentralized field? Doing your due diligence why did Loopring choose LINK. What advantages did Link have over said other oracles?
A: We looked at some other decentralized ones including those that are not necessarily oracle providers by design, but produce semi-reliable price feeds as a by-product, such as using the LRC/ETH price from Uniswap, Kyber, etc. That would have been possible. But as you know, would likely be more manipulatable than having an aggregated, redundant source such as Chainlink. (Which pulls from those sources perhaps, and others…). So we viewed Link as having a more robust feed that could defend against future attacks. We’ve gone through so much R&D and “trouble” to ensure Loopring v3 is absolutely the most secure in its design, vs other layer 2 scaling solutions available (i.e. we use ZK Rollup vs Optimistic, and on-chain data availability vs off-chain data), so we wanted this other ‘link’ in our system to stand up strong to potential attacks as well :).
Q: It seems that a trend popping up is to try to move as much of the computing process off-chain to improve scalability. Is there a point where this begins to hinder decentralization, or am I off base?
A: Hi Maple Syrup, good question, but it is slightly off base, at least in our case :). The off-chain computing we do is often referred to as ‘verifiable computation’ because it is correct by construction — its faithfulness guaranteed by Zero-Knowledge Proof cryptography. Without sounding too simplistic, it is math — the ‘state’ of anything is either provably true or false… no questions to its validity. So, decentralization is not sacrificed at all, because the validity itself is mathematic, and the proof is being verified by Ethereum ITSELF… so absolutely no compromise on layer 1 guarantees. But actually, there are competing off-chain constructions which do sacrifice decentralization, that is just not something we are willing to do :). As mentioned above, the big design choices are ZK Rollup vs Optimistic, and on-chain data availability vs off-chain data availability. We take the former in each, which means Ethereum-level security. If someone deviates from those choices, you are right — they are sacrificing decentralization to some degree, by relying on some combo of fewer nodes/people, cryptoeconomic security, etc. This short article may be very helpful to understand why Loopring v3 does not hinder decentralization: always choosing security over marginal scalability.
Q: With DEXes, will the creation of new financial instruments be as simple as writing up a smart contract?
A: I’ll keep this one short: Yes :). It’s actually even simpler than ‘writing up’ a smart contract. It can be as simple as calling one function on an existing smart contract…and boom, you have some new instrument, market, etc. What web2 did for information, connecting people, etc, Etheruem / web3 will do for finance and value. To be clear, these instruments will likely live outside a DEX (and be powered by protocols with a specific functionality — derivatives, loans, etc) and they will be traded on DEXes.
Q: Do you expect to use further aggregators and new pairs any time soon? What pairs? With which frequency can we expect job requests?
A: We do. Other pairs will be needed for the advanced order types (stop loss, etc) referenced in the video. Any asset / pair we want to endow with that capability will require a faithful price feed. Also, other price feeds will be needed for the smart contract wallet, to set allowances for certain user activity.
Q: How confident are you that Ethereum remains the primary smart contract platform for the foreseeable future?
A: I’ll keep this one short too: Very :). That is my personal opinion (and that of the Loopring team), but we feel that thesis getting stronger every day. I want expound too much, but I just find it will be very hard to beat the mix of: developers, head start, actual usage, composability, community, DeFi, ETH being a valuable base layer money, mindshare, ethos, pace of innovation, flexible yet focused monetary policy (minimum viable security), etc… Of course, cannot rest on its laurels, but I see zero signs of that.
Q: what does the LRC token do for Loopring? What do you see token doing going forward?
A: Hi Fastlane. The LRC token is a staking token, used in 3 different staking mechanisms (depending on whom is doing it, and their goals). Please read here for info, but briefly, they are:
- Anyone can stake LRC to earn part of 70% of the protocol fees of all exchanges built on top of Loopring. (20% will fund the Loopring DAO and the remaining 10% will be burned).
- An exchange owner needs to stake LRC for economic security & reputation.
- LRC can be staked to lower the protocol fees on a specific exchange by an exchange owner, market makers and high-frequency traders.
Chainlink of course is helping make LRC staking type 2 (DEX security bond) more robust with a reliable price feed.