Consensus for Networks That Keep Falling Apart

Almost every blockchain assumes the network works. Nodes can reach each other, messages arrive, partitions are temporary and rare. Proof-of-Gossip starts from the opposite assumption: the network is broken most of the time. It is a consensus protocol for disaster zones, remote areas, conflict regions, and IoT deployments that have no reliable internet, and its design follows from taking that constraint seriously.

Designing for the Partition, Not Around It

The interesting technical inversion here is that intermittent connectivity is treated as the normal case rather than a failure to recover from. Standard consensus wants synchronous agreement, which quietly requires that nodes can talk to each other when they need to. In a partitioned network that assumption never holds, so the protocol is built for eventual consistency and asynchronous consensus instead. Nodes make progress locally and reconcile when connectivity returns, rather than stalling until a quorum is reachable.

Connectivity itself is hybrid. When there is no internet, nodes communicate over LoRa mesh networks, which trade bandwidth for range and work without any infrastructure. When an internet gateway is available, the protocol uses it. That gives you a system that degrades to radio mesh under the worst conditions and takes advantage of real connectivity whenever it appears, rather than depending on either one.

Cheap Enough to Deploy Where It Is Needed

The hardware target is a real constraint that shapes the whole design. The protocol runs on $10 ESP32 devices. That price point is not a bragging metric, it is what makes the intended deployments possible at all: places where you cannot count on power, internet, or expensive equipment need nodes cheap enough to scatter widely and lose a few. Designing to fit an ESP32 forces discipline about memory and message size that a protocol assuming a datacenter never has to think about.

The documentation set reflects how much of this project is design rather than code so far. There is a complete whitepaper, the original consensus design based on Meshtastic, an eventual-consistency validation analysis studying feasibility, a hybrid-connectivity addendum for gateway and cross-region protocols, and a separate document for extreme network segregation in conflict zones. This is legitimate infrastructure research aimed at resilient communication where normal networks fail.

The Takeaway

The lesson I take from Proof-of-Gossip is that your hardest assumption is worth naming and then deleting. Most consensus work assumes reachable peers; deleting that assumption changes everything downstream, pushing you toward eventual consistency, mesh radio, and hardware cheap enough to deploy in the field. The project is currently in a pre-seed, pre-development stage, and the honest status is that this is still mostly whitepaper and design work rather than a running network. But the design question it starts from, how do you agree when you usually cannot talk, is a real one.