Early Days of Ethereum

Preserving the history and stories of the people who built Ethereum.

swarm

Swarm is the decentralized storage layer of the Web3 Holy Trinity, serving as the world computer's disk

Swarm (protocol name: BZZ) is the decentralized storage and content distribution layer of the Holy Trinity of Web3, alongside Ethereum (computation) and Whisper (messaging). In the "world computer" metaphor, if Ethereum is the CPU, Swarm is the disk.

Origins

The concepts behind Swarm date back to at least 2011, originating from Daniel Nagy's work on improving BitTorrent. Nagy had operated DC++ nodes and experienced legal troubles related to decentralized file sharing, which motivated his interest in building a more robust, privacy-preserving, and incentivized storage system.

BitTorrent had several limitations that Swarm aimed to address:

  • Slow startup times and throttling
  • Lack of proper incentives to keep data available
  • Individual file chunks not being directly addressable
  • Reliance on centralized trackers
  • Poor privacy and deniability properties

Technical Architecture

Daniel Nagy presented his vision for Swarm at DEVCON0 in Berlin (November 2014), introducing key concepts that would define the project.

The core innovation was the DPA (Distributed Preimage Archive) - the idea that instead of storing just the location of data in a distributed hash table (DHT), you store the actual data itself at the address determined by its content hash. This approach provides:

  • Content addressing - Every chunk is addressable by its hash
  • Automatic distribution - Popular content naturally spreads across the network
  • Plausible deniability - Nodes can't control what data lands in their "neighborhood"
  • Kademlia routing - Efficient lookup using logarithmic distance metrics

The SWAP protocol (Swarm Accounting Protocol) handles bandwidth incentives through peer-to-peer accounting. Nodes track what they forward for each other, and settlements occur when imbalances grow too large. This was one of the earliest practical implementations of bandwidth incentivization in a decentralized network.

Development History

Viktor Trón recalled in his Early Days of Ethereum interview that the Swarm team coalesced around Daniel Nagy after DEVCON0:

"I kind of gathered around him and continued on churning out some code… slowly, slowly kind of organically became a little bit of a sub team after that."

The core team included Viktor Trón, Daniel Nagy, and Zsolt Felföldi (who was originally hired to work on Swarm but was redirected to the light client by Jeff Wilcke). Nick Johnson also worked on Swarm and developed ENS while part of the team.

During the Ethereum Foundation's austerity period in 2015, Vitalik Buterin suggested potentially using IPFS instead and focusing on an incentive layer. However, Jeff Wilcke quietly allowed Swarm development to continue.

Key milestones:

  • December 2016 - First public pilot released as part of Geth 1.5
  • 2018 - Proof of concept 3 released; standalone Swarm client
  • 2019 - Swarm Summit in Madrid; began spinning out from EF
  • 2021 - BZZ token launched; Swarm Association established in Switzerland
  • 2021 - New Bee client released, switching from devp2p to libp2p

Comparison with Alternatives

Compared to IPFS/Filecoin and Arweave, Swarm offers:

  • Bandwidth incentives - SWAP protocol for real-time accounting
  • Privacy properties - Forwarding-based retrieval obscures request origins
  • Postage stamps - Pay-as-you-go storage rent model
  • No deal negotiation - Automatic pricing based on supply/demand

Primary Source

Much of this article is drawn from Viktor Trón's Early Days of Ethereum interview, which provides extensive first-hand accounts of Swarm's origins and development.

Resources