1 of 32

CONNEXTing the dot

11-12-2022

by chompk.eth

@cpkcpk3

2 of 32

List of Contents

🌉 Introduction to Bridges

🕓 History of Connext

⚖️ NXTPv1, Trustless Bridge

🐺 The Amarok Upgrade

🔑 Security of Amarok

💻 Technical Details

2

3 of 32

🌉 Introduction to Bridges

3

For Connext Network

4 of 32

🌉 What is Bridges?

4

Blockchain A

Blockchain B

Communication channel between two independent blockchains. The goal is to have a protocol that send messages between blockchains

Asset bridging

0x0Ab93eFD…

Message bridging

Bridge

What makes each bridge different despite having a similar functionality?

Verification Method

5 of 32

Categorizing Bridges Verification

Types of bridges verification:

5

Externally Verified

Natively Verified

Locally Verified

Example:

Multisig, PoS, LZ Oracle/Relayers

The bridge events data are verified by an external party. This bring a trust assumption on how external party relaying the data

The bridge events are natively verify using its own complex architecture. Communication are secured as long as both parties are in the same ecosystem

Example:

ETH<>L2*, Polkadot XCMP, Cosmos IBC

Data transfer are done via local communication between two individuals on both chains. The transfer protocol are economically adversarial, thus safe and secure

*Only ETH<>L2 communication, L2<>L2 is not natively verified

Example:

Hop, Connextv1

V1

6 of 32

Interoperability Trilemma

6

  • Arbitrary message passing
  • Difficult to expands to other domains outside the ecosystem
  • Bridge security doesn’t introduce new trust assumptions apart from default domain security

Generalizable

Extensible

Trustless

Don’t introduce any additional trust assumption. Trustless bridge should have equivalent security to the underlying domain

Easily expands to other domain with minimal development and deployment efforts

The bridge should be able to transfer any type of messages, not restricted to just assets with economic value

V1

  • Arbitrary message passing
  • Easy to expands to other domains
  • Bridge security relies on a trust on third party (multisig, PoS, Oracles, etc.)
  • Only assets with economic value can be bridged
  • Easy to expands to other domains
  • Bridge security doesn’t introduce new trust assumptions apart from default domain security due to its economically adversarial nature

Generalizable & Extensible

Trustless & Extensible

Trustless & Generalizable

7 of 32

Optimistic Bridges

Solving Interoperability Trilemma, introducing new tradeoffs

7

Optimistically Verified

Example:

Nomad Optimistic Bridge, Celo Optic Bridge

Transactions are submitted directly, but unfinalized. Watchers are incentivized to verify any malicious bridge events during a challenging period

watchers

⌛️

Extensible, Generalizable, Trust-minimized, High Latency

Generalizable

Extensible

Trustless

Low Latency

Optimistic Bridge introduce new tradeoffs on latency with minimal trust assumptions

8 of 32

Economic Securities of Bridges

Consider a verification systems that requires a total of N actors

8

Externally Verified

Optimistically Verified

Locally Verified

Natively Verified

It takes M out of N actors colluding to corrupt the system

It takes k out of N corrupts validators to corrupt the system

It takes the domain breach to corrupt the system

It takes all N actors colluding to corrupt the system

watchers

⌛️

Example of security breach:

  • Multisignature compromised up to certain threshold

Example of security breach:

  • No watchers report for malicious events
  • Updater fraud

Example of security breach:

  • The smart contracts of the blockchain itself is compromised

Example of security breach:

  • 51% attack

9 of 32

🕓 History of Connext

9

10 of 32

History of Connext

10

Connext was founded with the goal of scaling blockchains and bring blockchain to broader consumer public

2017

2018

May 2020

Dec 2020

Became one of the first layer-2 (L2) R&D Team. Built the first layer-2 on Ethereum called the State Channel Networks designed for payments. Later, the team have worked with other L2 teams such as OmiseGO team, Matic team

Built a very first Arbitrum-Optimism bridge called Spacefold as a POC. This is the very first trust-minimized cross-chain bridge for rollups.

Built a cross-chain transfer protocol called Vector. Collaborated with 1Hive and built a POC bridge that later become xpollinate

(The old name of Connext Bridge, read more here)

11 of 32

History of Connext

11

2021

2022

Launched NXTP, a locally verified liquidity protocol

Introducing the Amarok Upgrade 🐺, allowing cross-chain generalization and modular bridge design. In late 2022, the team launched beta launch of Amarok to the mainnet

12 of 32

⚖️ NXTPv1, Trustless Bridge

12

13 of 32

NXTPv1, Connext before Amarok 🐺

A trustless, locally verified bridge

13

Chain A

Chain B

1. Route Selection

2. Prepare

3. Fullfill

I want to transfer 1000 DAI from Chain A to Chain B

Routers bid offers using their liquidity provided on both Chain A and B

Routers

Chain A

Chain B

Routers

Router that win bid prepare liquidity on Chain B

User sign transaction to send 1000 DAI from Chain A to Chain B

🔒

🔒

Chain A

Chain B

Routers

🔓

🔓

🔑

User sign a message to unlock the funds on both side. Completing the bridge transaction

Users select the chains and assets and start an auction with the transfer information, and routers bid to complete the transfer with a time/price range

Users send a transaction to the nxtp contracts on the sending chain. Upon seeing the contract event, routers lock up liquidity on the receiving chain

Users provide a signature to unlock funds on the receiving chain, and routers use the revealed signature to claim the liquidity on the sending chain.

14 of 32

Problems with NXTPv1

14

Problems for End Users

Problems for Developers

Problems for Routers

⛽️ Gas Costs

🖋 Signing to Claim

🔐Fund Lockup Risk

⚡️ Speed

⛲️Liquidity Fragmentation

🛎 Offchain Auction

🖋 Signing to Claim

🤖 No Generalized Messaging

🔄 Rebalancing

💸 Unclear ROI

📜 Strict Liveliness

💩 Gas Griefing

Two phase txs (prepare/fullfill) is costly and hard to batch

Signing to release locked funds is bad in term of UX

Poor routers liveliness can freeze users’ funds for up to 72 hours

Poor routers liveliness can also affect transaction speed

Only one router can fullfill a transaction, low capital efficiency

Developer requires to run a client-side SDK for routers auction

Signing to claim adds complexity when design user journey

NXTPv1 allow cross-chain smart contract call only for a case that there’s economical value. Causing a bad developer UX.

Two-side liquidity requires asset rebalancing, reducing capital efficiency

ROI is hard to track with two-step transaction

Poor router liveliness cause users’ funds to be stuck. Making running routers difficult

Canceled transaction doesn’t provide reimbursement for users

15 of 32

🐺 The Amarok Upgrade

15

16 of 32

🐺 Amarok Upgrade

🦾 Solving Problems from NXTPv1

  1. User Experiences
  2. 🖱 1 click UX
  3. Router Experiences
  4. ⛲️ Better liquidity, No rebalancing
  5. 🎛 1:N Routing instead of 1:1
  6. Modular Design

🔧 Introduce New Functionality

    • 🤖Arbitrary Message Passing
    • 🤑Cheaper & Faster txn

How? Let’s Explore

16

Connext enable interoperability to rollups, L1 using a Hub-and-Spoke model

17 of 32

Modular Bridge Design

17

Transport

Execution

Liquidity

Getting the right asset for users

This is handle by Connext AMM

Transport data from source_chain to dest_chain

Handle by off-chain infrastructure

Unpack and execute txn

This is handle by Connext’s backend

Verification

Data verification, ensure cross-chain data integrity

Can be varied: AMBs for L2, IBC for Cosmos,

Optimistic Bridge by Nomad, etc.

Monolithic Bridge

Monolithic bridges handle all functionalities by themselves

Connext’s modular bridge decompose functionalities, allowing easy composability

18 of 32

How it works?

18

Hub

Spoke

Bypass slow path with router’s fast liquidity

Fast path

Transaction fulfilled by off-chain agents (routers, relayers, sequencer)

Slow path

Executing the transaction on the verifier layer (AMBs, IBC, XCMP, etc.)

Verification Channel

(AMBs, IBC, XCMP, etc.)

Dest Chain

Smart Contract

xApps

Smart Contract

xcall

callbacks

Source Chain Smart Contract

19 of 32

A high-level Connext Transaction Flow

19

Underlying bridges

(AMBs, IBC, XCMP, Optimistic Bridge, etc.)

Slow messaging

Routers: Fast Liquidity

AMM

AMM

source_chain

dest_chain

Passive LPs

Active LPs

Passive LPs

DAI

NextDAI

NextDAI

DAI

20 of 32

Off-chain Actors

20

Routers

Sequencer

Service by Gelato

Relayers

Destination Chain

Smart Contracts

$

$

$

$

$

Decentralized network that execute smart contract transactions for users in exchange for a small fees

Collects bids from all chains and randomly select router(s) to fulfill them

Liquidity providers, enable instant liquidity for users. Also simulate a fast path transaction

Anybody can participate, no minimum liquidity required

21 of 32

A Detailed Connext Transaction Lifecycle

21

Bid txn

Transport

Execution

Liquidity

Verification

Hub

source_chain

dest_chain

Spoke

Spoke

Hub

Off-chain Agents

Off-chain Agents

Application

Smart Contracts

Smart Contracts

xApps Smart Contract

AMMs

Getting right assets for AMBs

*AMBs

txn event

Connext

Contract

Call AMBs and emit txn event

Routers observe txn event

xcall

*AMBs

Sequencer

Relayer

Routers

Routers

Simulate observed txn

xApps Smart Contract

AMMs

Catch bid txn

Get right assets from AMBS

Relayer execute smart contracts

Hashed params

Store

Slow *AMB message

goes through Spoke via Ethereum

Validate

Return funds to Routers

xreceive

*This component can be interchangeable, depends on underlying verification protocol (e.g. XCMP, IBC, etc.)

22 of 32

Connext

Contract

Spoke

Connector

MERKLE

Connext

Contract

Hub

Connext

Contract

Spoke

Connector

MERKLE

MERKLE

Routers

AMB

AMB

23 of 32

🔑 Security of Amarok

23

24 of 32

Economic Security of Connext

Depends on the verification protocol that Connext used, different protocol can have different economic security

24

Transport

Execution

Liquidity

Verification

AMBs - L2 to L2 Communication

Optimistic Bridges - General Cross-chain Communication

XCMP, IBC - For communication in Cosmos/Polkadot ecosystem

AMBs (Arbitrary Message Bridges), or a default ETH to Layer-2 bridges, is a natively verified bridge. Therefore, any L2 to L2 communication that utilized AMB as a verification layer have the same economic security as Ethereum mainnet

Optimistic bridge economic security depends on the number of watchers monitoring the bridge events. The cost to corrupt the system is equal to corrupting the whole watchers, which is very expensive

XCMP and IBC is also a natively verified bridge protocol. It have same economic security as their underlying domain

25 of 32

Risks of Amarok, and its Trust Assumption

What do we trust when using Connext?

25

Sequencer

Underlying Verification Security

The sequencer are capable of censorship and extracting MEV, but the transactions are fully secure and trustless. This problem can be mitigated by having a decentralized sequencer

The protocol relies on the underlying verification security such as AMBs, Optimistic Bridges, XCMP, IBC, etc. When using Connext, you’re trusting that the verification layer is safe economically

`

Verification

26 of 32

What’s $NEXT

Governance system for Connext Infrastructure

Powered by $NEXT Token

  • 1B Total Supply
  • ERC-20
  • Distribution through contributor program

26

Let’s DAO 🏛

Phase 1 : Completed

Mainnet Launch

🚀

Phase 2 : Soon ser

Token

Launch

We are here! Testing Amarok upgrade on Mainnet!

Connext will pivoted to DAO structure powered by $NEXT

27 of 32

💻 Technical Details

27

28 of 32

EIP-2535: Diamonds, Multi-Facet Proxy

28

Immutable

Smart Contracts

Proxy Smart Contracts*

Multi-Facet Proxy Smart Contracts**

Immutable, unupgradable, limited contract size

Upgradable contracts using proxy pattern

Modular design pattern, multiple proxy to each contracts with each module

Implementation Contract

Proxy Contract

Delegate call

Delegate call

Module Contracts

(Facet)

Proxy Contract

(Diamond)

Smart Contract

29 of 32

💎 Connext Diamond

29

delegate call

0x8898B472C54c31894e3B9bb83cEA802a5d0e63C6

Connext Diamond

Bridge Facet

Router Facet

Relayer Facet

Stable Swap Facet

Token Facet

0xe37d4F73ef1C85dEf2174A394f17Ac65DD3cBB81

0x0e93f1e9a44AF236186a8cBCdeb22e2E6564856c

0xBe8D8Ac9a44fBa6cb7A7E02c1E6576E06C7da72D

0xcCb64fDf1c0Cc1aac1C39E5968E82f89c1B8C769

0x9AB5F562Dc2aCcCd1b80d6564B770786e38f0686

Handle xcall, add/remove sequencers

Handle routers functionality (e.g. initialize, add, remove, etc.)

Handle relayers related functions

Handle liquidity swap for local assets

Handle Connext-based asset (e.g. NextDAI, NextUSDC, etc.)

30 of 32

Is Diamond Pattern an implementation risk?

30

NO.

Contracts are upgradable, does it means that Connext have a total control of pushing malicious code?

Connext smart contracts are:

  • 5/7 multisig (3 from Labs, 4 from foundation, external protocol devs, investors)
  • 1 week timelock on any upgrades

After the token launched, the control will be given to Connext DAO, and timelock would still remains

🐺 Amarok

Connextv1

NO.

Connextv1 use a NO upgradability patterns. This means that the codes are immutable

But this prevent the teams to fix/improve the protocol which is crucial

Immutable

Contracts

31 of 32

📚 Reference

31

32 of 32

References

Special Thanks to

32

sanchaymittal#8853 Engineer @ Connext

arjun#0082 Founders @ Connext