Audit Results & Launch Plan
To reduce the surface area that we need to iterate on, we’ll be rolling out the network in stages. Stage 0 will go live next Monday with a fully-noncustodial MVP network. Stage 1 will add support for router-router interactions. Stage 2 will remove reliance on all centralized services.
Six weeks ago, we announced the testnet launch of ↗️ Vector, our cross-chain/L2 routing network. Today, we’re excited to publish our audit results and share some more information on how we expect to roll out the full network!
🔍 Vector Smart Contract Audit Results
Note that the above audit only encompasses onchain contract code and the ways in which the protocol interacts with the solidity contracts. Offchain protocol code was out of scope for this review.
We’re fairly confident in the security of the offchain code. However, we still fully intend to do a more comprehensive audit of the offchain protocol structure early next year. Since the offchain framework typically takes longer to review (2–3 months), we’ve elected to do it in parallel to the rest of the network rollout.
📋 Network Rollout Plan
While speaking to projects in the space about Vector, it’s become extremely clear that this technology needs to be live and functional on mainnet ASAP.
Because of this, we decided early on to split the rollout of the network into a few stages, starting with a limited (but fully non-custodial) “MVP” release and progressively decentralizing and adding functionality until the full network is live.
Stage 0: Minimum Viable Network
Stage 0 will go live Jan 4th 2021.
Stage 0 will allow peers in the network to transfer value across chains noncustodially through one intermediary router. Routers can be operated by anyone and can open channels to any end user, but router to router interactions will not yet be possible. This means that a dApp wishing to enable onboarding from Ethereum to Matic would need to find a router that supports those specific chains. We will also be using a centralized messaging system while we explore Typescript implementations for existing p2p networking stacks like Libp2p and Matrix.
Stage 1: Multihop
Stage 1 will go live mid-late Q1 2021.
Stage 1 will enable router-router interactions allowing transfers to hop through multiple intermediary routers prior to reaching their destination. Importantly, this will mean that any user will be able to transfer to any chain, regardless of what router they’re connected to, reducing the reliance on a fixed set of known router operators. To bootstrap, we’ll initially use a centralized service for new router discovery, validating new channels in the network, and keeping track of network topology.
Stage 2: P2P Networking
Stage 2 will go live in Q2 2021.
Stage 2 will remove reliance on the above centralized services. We’ll migrate to a p2p networking stack (or roll our own if nothing robust is available) and move network topology data storage/validation to the routers themselves. Additionally, we’ll roll out a novel auction-based mechanism for router discovery which will remove the need for users to consciously find and connect to public routers.
⚖️ Future Work
At the moment, the Connext team is laser focused on getting to a fully decentralized network by mid next year. In parallel, however, we’ve been keeping an eye on the governance space, with the intent of handing over the ongoing maintenance and development of the protocol entirely to the community. We’ll keep you updated as we learn more here!
❓ Have any other questions?
If you have any questions, get in touch with us by:
And lastly, please share this with your friends in the ecosystem if you find it interesting.
Thanks for your support,