DFINITY founder Dominic Williams has shared the story of a failed cloud-hosted game that he says shaped his thinking about decentralisation and influenced the design of the Internet Computer.
Writing on X, Williams described what happened to Fight My Monster, an online game and social network he founded before moving deeper into the early cryptocurrency ecosystem. The game had reached about three million users and was financially self-sustaining, according to Williams.
Fight My Monster was built between 2010 and 2012 around a horizontally scalable architecture designed to handle growing numbers of users without requiring major changes to its backend software.
Williams said the system had three main components: web hosting for the game’s frontend and assets, a game server framework called Starburst, and the Cassandra NoSQL database.
The architecture allowed the game to expand by adding hardware as demand increased. Cassandra, in particular, enabled storage, read and write capacity to be scaled by adding more nodes.
Williams said the database used a replication factor of five, meaning multiple copies of data were maintained across different nodes. He described the system as fault tolerant, although it was not Byzantine Fault Tolerant like a blockchain.
The infrastructure still required some maintenance. Database nodes would occasionally fail under the volume of data being generated, requiring Williams to remove failed machines and add replacements.
Despite those challenges, he said the architecture allowed Fight My Monster to grow with relatively little administration.
The problem came from somewhere outside the technical architecture.
While travelling in Asia and attending events connected to the early Ethereum community, Williams said he stopped checking his email regularly. During that period, the credit card being used to pay for the game’s Cassandra nodes expired.
According to Williams, the cloud provider sent several warnings but received no response. After a few weeks, he said, the provider deleted the nodes.
When Williams returned to Palo Alto, California, he discovered that the data loss could not be reversed.
“Fight My Monster was dead,” he wrote.
The incident eventually led to a broader lesson for Williams about decentralised infrastructure. He argued that an infrastructure system can be highly scalable and fault tolerant while still having a single point of failure that can bring the entire service down.
He said the experience became part of his thinking around the design of the Internet Computer, particularly the importance of looking beyond the number of nodes in a network.
Williams argued that node counts alone do not necessarily demonstrate meaningful decentralisation. If many nodes are controlled by a small number of independent operators, or hosted through the same cloud providers or data centres, a network may have less operational independence than its raw node count suggests.
He described the number of independent parties operating nodes, along with their physical and jurisdictional distribution, as important considerations in assessing decentralisation.
Williams said this thinking contributed to the concept of “deterministic decentralisation”, which he said will feature in a new form through ICP cloud engines.
According to his post, ICP cloud engines are designed to allow enterprises to create an Internet Computer subnet whose configuration they control by selecting nodes from a marketplace. Williams said more than 1,500 nodes are currently waiting in the pool.
The cloud engine model is intended to provide a serverless cloud environment where applications can run with properties including tamper resistance, continuous availability and support for digital assets.
Williams also highlighted the ability to change the nodes supporting a cloud engine without interrupting the applications running on it. This includes changing the replication factor or replacing nodes, which he said can reduce dependence on a single compute provider.
The system can also support different geographic and infrastructure requirements. Williams gave the example of a German enterprise selecting nodes located within the European Union to address requirements such as GDPR compliance.
The cloud engine console is served from the Internet Computer network and updated through the Network Nervous System. Williams said it can warn users when they select nodes that share a provider, data centre or cloud operator, encouraging greater infrastructure independence.
For Williams, the experience of losing Fight My Monster remains relevant to the way decentralised networks are built.
The former game may have disappeared after its cloud infrastructure was deleted, but the lessons from that failure have continued into his work on decentralised compute.
Williams ended his post by saying he looks forward to sharing ICP cloud engines more widely.
Dear Reader,
Ledger Life is an independent platform dedicated to covering the Internet Computer (ICP) ecosystem and beyond. We focus on real stories, builder updates, project launches, and the quiet innovations that often get missed.
We’re not backed by sponsors. We rely on readers like you.
If you find value in what we publish—whether it’s deep dives into dApps, explainers on decentralised tech, or just keeping track of what’s moving in Web3—please consider making a donation. It helps us cover costs, stay consistent, and remain truly independent.
Your support goes a long way.
🧠 ICP Principal: ins6i-d53ug-zxmgh-qvum3-r3pvl-ufcvu-bdyon-ovzdy-d26k3-lgq2v-3qe
🧾 ICP Address: f8deb966878f8b83204b251d5d799e0345ea72b8e62e8cf9da8d8830e1b3b05f
Every contribution helps keep the lights on, the stories flowing, and the crypto clutter out.
Thank you for reading, sharing, and being part of this experiment in decentralised media.
—Team Ledger Life
Community Discussion