Topology Protocol
Advancing humanity’s presence in cyberspace
Founded in late 2021, Topology’s founding vision was to create sovereign, persistent, and interoperable realities by advancing the composability and interoperability properties of onchain programs. We conducted a series of experimentation: Isaac explored multiplayer construction mechanics on top of numerical simulations of multi-body systems; MuMu explored Zachtronics-inspired game loops where players solve visual programming puzzles; Shoshin explored the behavioral complexity of user-created agents situated in a physics simulation environment running in Cairo zkVM.
Blockchain by design has plenty of core limitations: it is single-threaded, charges every transaction with fees, and forces programs to interact in a strongly coupled manner. The modular thesis advocates for parallel application threads settling their execution traces in compressed form on the state of a shared parent thread for security. However, each application thread is still a single thread. Advocates for verifiable computing proposes transporting the execution traces of application threads to parent threads in proof-carry data, justifying the use of centralized high-performance sequencer. This trade-off reintroduces centralized trust assumptions on the application layer and creates bottleneck in liveness. We refuse to believe that the upperbound of decentralized applications lies on this Pareto curve.
In early 2023, we discovered the multiple lines of research pursuing coordination avoidance in distributed systems. Byzantine consensus allows distributed agreement in total ordering of atomic operations in the presence of Sybil actors. This model preserves strict serializable consistency, allowing for the distributed execution of Turing-complete programs. There exists a class of programs — invariant-confluent programs [1] — that can execute correctly in distributed fashion without coordination, which means no voting or “consensus” is required to reach consistency. By using hash graphs to encode causality information in a self-certifying manner, and by using suitable causal broadcast algorithm for information dissemination, invariant-confluent programs can execute in open peer-to-peer networks with immunity to Sybil attacks without relying on any proof of resource or honest majority assumption [2]. This enables decentralized applications to have zero-fee transactions and local speed (lock-free) progression. For programmers, Conflict-free Replicated Data Type (CRDT) [3] conveniently abstracts lock-free replication strategies and expose the APIs of familiar abstract data types. CRDTs can compose into richer CRDTs, providing the affordance of class composition. BFT-CRDT [4] provides causal ordering guarantees in Byzantine environment. We conceptualized live objects, or BFT-CRDT actors that interact in asynchronous message-passing [5].
We believe live objects provide fundamentally new affordances to the developers of decentralized applications. The affordances of zero-fee, local speed, loose coupling and Sybil immunity complement the Byzantine consensus guarantees of blockchains. Also, client-server applications large and small could interoperate through live objects just as they already do through smart contracts. Our founding vision remains, only that our approach has crystalized: the credible neutrality, high concurrency, credible neutrality, and open-world programmability requirements of the open metaverse and autonomous worlds call for core innovation in the architecture of the world computer, a mission originated from Ethereum. We see ourselves as world computer researchers and engineers who continue that mission.
Follow Topology’s website, X, Farcaster, and Discord for future updates. Join us in our shared quest to advance humanity’s presence in cyberspace.
Reference
[1] Bailis et. al., Coordination Avoidance in Database Systems, 2014.
[2] Kleppmann et. al., Byzantine Eventual Consistency and the Fundamental Limits of Peer-to-Peer Databases, 2020.
[3] Shapiro et. al., Conflict-free Replicated Data Types, 2011.
[4] Kleppmann, Making CRDTs Byzantine Fault Tolerant, 2022.
[5] Hsueh et. al., Topology Protocol: BFT-CRDT actors for highly available decentralized applications (DRAFT), 2024. https://github.com/topology-foundation/paper