Bitcoin—now fifteen years old—put blockchain on the map. Today, the technology is deemed essential to the next incarnation of the internet. Web 3.0 will be blockchain-based. Tomorrow’s decentralized web will operate without a central authority. But first, blockchain stakeholders will have to innovate to address several challenges:
The purpose of the Toposware project is to find solutions to these issues. And the CEA is contributing its knowledge of distributed systems. CEA-List has a strong track record with several partnerships and publications around blockchain. The institute has worked on the Tendermint blockchain and helped draw up specifications for Tenderbake, the protocol used by Tezos.
Toposware and CEA-List will be focusing on cross-chain communication systems. Together, they will tackle the challenge of making sure messages sent by different blockchains are “seen” in the correct order. Currently, messages can be seen in different orders by different blockchains—something that is problematic for financial transactions.
One solution is to use a central blockchain to set up communication between different blockchains. Another is for the participants in the sending and receiving blockchains to handle communication, in which case the blockchains must give up the privacy of their data. The Toposware project will result in a new cross-chain communication protocol that does not require a central authority. The project partners have opted for a causal order broadcast protocol that maintains the order of the messages sent.
CEA-list and Toposware also investigated how to best implement incentives for users to ensure that the protocol works optimally—with no cheating. A blockchain is a self-managed system. So, users can do whatever they want, at least in theory. Spending cryptocurrency twice is a well-known example. Because there is no physical money, the same coin could theoretically be sold twice to two different people. But there are protocols designed to prevent this kind of fraud, usually by making “bad” behaviors either very complicated or in some way unfavorable to the actor and by encouraging, through incentives, “good” behaviors. Applying an incentive mechanism to a causal order broadcast protocol could effectively ensure that, when a blockchain sends a cross-chain transaction, it is duplicated properly across all the nodes in the system.
The project, which kicked off in October 2021, could start producing results by the fall of 2022. The decentralized finance, or DeFi, industry, which uses multiple blockchains, could be among the first to benefit from the advances made. The new protocol could also allow DApps, the decentralized apps deployed on the blockchain, to be reused on multiple blockchains, for instance. More broadly, the project will remove some of the hurdles to the widespread adoption of blockchain technology.
CEA-List’s knowledge of distributed systems was a crucial addition to our own areas of expertise, cryptography and blockchain.
We have already put out a white paper. We are also planning several other publications, co-authored with the CEA, on incentive mechanisms and Sybil attacks.
Topos, our cross-blockchain solution, wasn’t scalable. Our partnership with CEA-List produced the primitive that will allow us to scale.