owned this note
owned this note
Published
Linked with GitHub
# transfers.wtf
This is an unconference style event. Please feel free to add a topic you'd like to talk about in Unscheduled Proposals below!
![image](https://hackmd.io/_uploads/r19rrELVa.png)
# Post Event Summary
*This will also be emailed to those who registered*
Konrad and Hugo from Peanut Protocol explained links can abstract away complexities around multichain and multi-account use cases. Specifically, they argued that the normal user flows for payments involves a long chain of offchain coordination messages about preferred tokens, chains and addresses. Here is an example of such a chain:
> Alice: I want to pay you $100, is usdc ok?
> Bob: I'd prefer dai. My address is bob.eth
> Alice: OK, I'll swap. Is Arbitrum ok?
> Bob: No. I'd rather have it on Optimism.
> Alice: OK! Here is the optimistic.etherscan link.
> Bob: Thanks!
Instead, these instructions can be (a) encapsulated in a simple standard data format, (b) made by the claimer via an escrow contract. Then Konrad and Hugo, presented how Peanut Protocol works: the sender deposits funds into an escrow onchain, but sends the key that is needed to withdraw from the escrow offchain, e.g. via a message. Then, Peanut Protocol presented their new cross-chain feature available on experimental.peanut.to/
Fig from Squid explained how Squid manages to go from intent to realisation in under 20 seconds. Squid provides an API and front end dev-tools to build cross-chain swaps, bridges and generic transactions. Squid routes using Axelar and always routes through USDC, which means incresed liquidity on both token pairs as TokenA/USDC and TokenB/USDC will have more liquidity than TokenA/TokenB. The way they can be faster than finality is by using boost. Boost is a smart contract overlay which allows a provider to optimistically fulfill a transaction immediately after the user completes their transaction on the source chain. The provider is providing a short term loan, which is guaranteed by the pending bridge transaction.
Markus from Propellerheads explained the P2P trading promise of intents: everyone is a DEX! When trading on decentralized exchanges (DEXs), you often don't receive market price due to DEX fees, gas costs, and price impact, which can affect both large and small trades. However, there's a way to mitigate these costs and achieve a price closer to the market rate. This involves using patient trading strategies. If you're willing to wait, you can engage in peer-to-peer (P2P) trades without the usual price impact. Additionally, this approach allows others to cover your gas expenses and can result in lower DEX fees. Essentially, patience in trading can lead to more cost-effective transactions. We went into more detail about:
* DEX fees: DEX pool fees 0.3%.
* Price-impact: a 100k trade on a token with 5m liquidity: you’ll easily pay 10% above market price.
* Gas: Gas can be $50 on a swap – this is 5%(!) on a 1k trade.
* Market lag: DEX prices randomly fluctuate 1% above and below market price because CEX-DEX arbitrage takes time.
This leads us to a future where offchain p2p matching will be the most efficient way of doing trades, leading to the eradication of DEX fees and gas fees, and reducing price-impact.
Here is a summary from the workgroups:
In the workgroup on cross-chain swaps, some interesting points have been made:
- ZKPs will greatly decrease the cost of bridging
- Many protocols are heavily centralised, e.g. when Multichain's core dev was arrested, they lost access to their multisig
- UX insights were around displaying assets as one total, the hierarchy being: account -> assets -> chain, rather than the current account -> chain -> account model. Some arguments were also made in favour of an assets-centric model that works with in a multi-account framework.
On the payments workgroup, the following points have been made:
- VISA charges 2% but it's way more complicated than an insurance fee
- VISA reduces risk by getting a lot of knowledge from the layer on top of VISA, e.g. checkout.com.
- Regulatory capture allows for high profit margins.
- Can we unbundle payments in crypto and get rid of the (unjustified) 2% fee?
- Yes.
The unscheduled presentations were by DC from Daimo.xyz. Daimo is a user-friendly wallet that allows simple Base payments with USDC. They heavily rely on links as a UX-friendly abstraction. Daim uses an ERC-4337 contract account, which brings a new level of security to blockchain transactions. Each device linked to an account that holds a secret key for authentication via FaceID or similar. Daimo, a non-custodial wallet, ensures that the keys stay within secure hardware like the iPhone's Secure Enclave. They have a naming service and a claim link feature for easy sending, making it a Venmo-like experience.
Another presentation was by GhostPay who used Railgun and Peanut for private transactions. The railgun zk-shielded account is the sender, so Alice (public) creates a zkAlice (private) and deposits into funds into Peanut Protocol from zkAlice. The recipient, Bob, can claim with his public account and only reveal himself to Alice, and no one else. Alternatively, Bob can also create a zk-shielded address and claim with that fully privately.
# Agenda
November 19th, 2023, DevConnect
| Time | Event |
| -------- | ----------------------------------------- |
| 4:00 pm | Tea & Drying |
| 4:30 pm | Multichain Nightmare (Konrad & Hugo, Cofounders at Peanut Protocol) |
| 5:00 pm | EZ Payments Workgroups |
| 5:30 pm | Bridging the Gap Between Intents and Their Realization (Fig, cofounder of Squid) |
| 6:00 pm | Break / Informal Discussions |
| 6:30 pm | Intents (Markus, cofounder of Propeller Heads) |
| 7:00 pm | Break / Informal Discussions |
| 7:30 pm | Mingling |
| 8:00 pm | End |
# Talk Slides
Peanut: https://docs.google.com/presentation/d/1QyJ1x9LKiaX-fMgl5c6-dVAYJDZu2sIoUzTBJl7-tfg/edit?usp=sharing
PropellerHeads: https://link.excalidraw.com/l/A4KtFovP5QQ/1FRQUv8CYKZ
Squid: https://docs.google.com/presentation/d/194GlKYiBP9ehXTAz_8m1a5S17QfrBlwIdBbExzw5p5I/edit#slide=id.p
## Unscheduled Proposals
- Daimo.xyz 10mins ACCEPTED
- GhostPay 10mins ACCEPTED
# Community Notes
## UX Workgroup
- Questions: reading state cross rollups
- the basis is the cross-chain messaging mechanism. currently, Squid uses Axelar. In the future, there will be a mesh of ZKPs. Axelar has the ability to plug in ZKPs. Li.Fi will just aggregate, so the ZKP question is not their direct concern. Squid's thesis is that there will be this fragmentation.
- Security assumptions. Li.Fi does a lot of security research. Uniswap Labs has a bridge assessment, who had some governance on one chain but the token on another chain.
- Multichain's dev was arrested and the whole protocol stopped.
- Hierarchy of ecosystem, account, wallet, token
- People mess up USDC and USDbC.
- Wishes: total value shown without worry about which assets are for which apps.
- USDC?
- USDT and Tron is commonly used IRL.
- What is the easiest onramp USDT Tron? It seems
- Decentralised onramping through ZKPs and zkemail.
- Instant PMF startup: zkemail fork on Tron USDT
- Binance P2P is quite decentralised
## Bridging the Gap Between Intents and Their Realization -
by Fig - Cofounder of Squid
https://docs.google.com/presentation/d/194GlKYiBP9ehXTAz_8m1a5S17QfrBlwIdBbExzw5p5I/edit#slide=id.p
- What are intents?
- Intents are a buzzword
- Intents vs Transactions
- What do people want with blockchain?
- NFTs, investing, etc
- Without Squid with many many steps
- Boost feature reduces to 20 seconds because there is no need to wait for finality (see below)
- Naive vs Non-naive model
- Naive intents model posits that intents can do anything
- The only thing that transactions only cannot do is speed
- transactions only can be: safe, let you buy things, be one-click, any chain/asset/app, show proof of purchase, reliability.
- transactions can always pre-define minimum gas costs and get the best prices
- Squid Boost
- The reason bridging is slow is because of waiting for finality on the source chain
- Boost fronts the funds and then does a refund
- Boost provider is currently permissioned
- Optimistic settlement saves a lot of time
- Use cases
- Easy buying in any token because it's fast
- Questions about gas fees
- Layerzero is cheaper because there are fewer notes (approx 3) whereas Axelar has 75 to verify the signatures
- UX insights from doing a type/voice-activated swapping: no one wants this
# Unscheduled Presentations
## Ghostpay slides
slide
- https://www.canva.com/design/DAFz25Kt1Ls/Dz8RobUPtluzrpm3GPFK6g/edit?utm_content=DAFz25Kt1Ls&utm_campaign=designshare&utm_medium=link2&utm_source=sharebutton
github
- https://github.com/porco-rosso-j/eth-rome-2023/blob/dev/README.md#demo-app
## How to get Lightning Fast - Daimo
by Dan Clements aka DC Posch, cofounder of Daimo
- Universal links that let payments happen
- Problem statements:
- You receive a link to get paid
- Link opens Chrome, opens Rainbow, go to Chrome, open Rainbow, open Chrome
- One-tap links are the future
- Shared domain, e.g. eth.limo, but also project domains (e.g. peanut.to, daimo.xyz)
- Deep links can be resolved by any address
- Demo of the product
- Backup mechanism
- It's a 4337 wallet
- It supports device keys and passkeys
- iOS has a secure enclave
- Secure enclave is like a ledger but it relies on device-native auth
- There are two elliptic curves, where Ethereum EOAs and Bitcoin use one type, but the rest of the cryptography world (e.g. bitwarden) use another system.
# Worgroups - Community Notes
## Intents & Payments Workgroup
- Payments
- Question: VISA has 2% chargeback fee, which works as an insurance fee.
- VISA's 2% does not come from the insurance. VISA reduces risk by getting a lot of knowledge from the layer on top of VISA.
- For example, checkout.com is the basis for Stripe. The infra is VISA / MasterCard
- VISA tells checkout.com tells that
- This trickles up percentages
- THOUGHT: actually , there's almost no cost to visa for chargebacks, because most txs are reversible
- money is actually only lost when it goes cross ecosystem, or cash is taken out
- TL: chargebacks were a big problem in crypto NFT marketplace. Ppl would buy NFTs with VISA card, but then run away with the nft and make a chargeback
- What they did to deal with chargebacks: KYC - they'd only accept payments from users that fulfill certain criteria (3Dsecure, etc)
- 3Dsecure is actually set by whomever is at the level of stripe (above checkout and above visa). They set it
- Regulatory capture
-
- Can we unbundle payments in crypto and get rid of the (unjustified) 2% fee?
- Credit Scores are very biased - but scores can be useful for merchants /
- Predatory financial system
- 15yo kids in the US get credit card in the mail with 500$, just like single moms in shelters get 4-5 credit cards through email
- Intents
- Cross-chain
## Partners Links
- Peanut Protocol: https://peanut.to
- Squid https://www.squidrouter.com/
- Propeller heads https://www.propellerheads.xyz/