The Network As
One Computer
The Network As One Computer
The Network As
One Computer
A 3-year research effort exploring whether distributed computation, storage, and coordination could be unified through a single consistent type system.
A 3-year research effort exploring whether distributed computation, storage, and coordination could be unified through a single consistent type system.
A 3-year research effort exploring whether distributed computation, storage, and coordination could be unified through a single consistent type system.
(
Mission
)
(
Mission
)
(
Mission
)
2019
We set out to build a new kind of development environment.
We set out to build a new kind of development environment.
Developers could describe what they wanted in plain English. AI agents would generate the code and infrastructure. Version control, build systems, and orchestration would speak the same language.
Developers could describe what they wanted in plain English. AI agents would generate the code and infrastructure. Version control, build systems, and orchestration would speak the same language.
We called it Console One. A unified layer for software development.
We called it Console One. A unified layer for software development.
(
Problem
)
(
Problem
)
(
Problem
)
Fragmented by Default
Building distributed software today means connecting a lot of separate systems - databases, queues, APIs, deployment tools, orchestration layers. Each one has its own logic, its own state, its own way of failing.
Building distributed software today means connecting a lot of separate systems - databases, queues, APIs, deployment tools, orchestration layers. Each one has its own logic, its own state, its own way of failing.
We wanted to see if there was a simpler foundation. One structure that could handle storage, compute, and coordination together — where every change is tracked and the whole system stays in sync.
We wanted to see if there was a simpler foundation. One structure that could handle storage, compute, and coordination together — where every change is tracked and the whole system stays in sync.
(
How It Works
)
(
How It Works
)
(
How It Works
)
The Model
1
One structure
Data, logic, and permissions all live in the same tree. Everything is defined in one place — if it doesn't fit the schema, it doesn't get in.
1
Freelance Digital Designer
Data, logic, and permissions all live in the same tree. Everything is defined in one place — if it doesn't fit the schema, it doesn't get in.
1
One structure
Data, logic, and permissions all live in the same tree. Everything is defined in one place — if it doesn't fit the schema, it doesn't get in.
2
Every change is a commit
Nothing happens without a record. Writes, reads, computations — all logged. Any state can be replayed. Any change can be rolled back.
2
Every change is a commit
Nothing happens without a record. Writes, reads, computations — all logged. Any state can be replayed. Any change can be rolled back.
2
Every change is a commit
Nothing happens without a record. Writes, reads, computations — all logged. Any state can be replayed. Any change can be rolled back.
3
No central server
Nodes own parts of the tree and stay in sync. Any machine can join, catch up on history, and participate without asking permission.
3
No central server
Nodes own parts of the tree and stay in sync. Any machine can join, catch up on history, and participate without asking permission.
3
No central server
Nodes own parts of the tree and stay in sync. Any machine can join, catch up on history, and participate without asking permission.
4
Security is structural
Access control isn't bolted on — it's part of the tree itself. If you don't have the path to something, you can't reach it.
4
Security is structural
Access control isn't bolted on — it's part of the tree itself. If you don't have the path to something, you can't reach it.
4
Security is structural
Access control isn't bolted on — it's part of the tree itself. If you don't have the path to something, you can't reach it.
(
status
)
(
status
)
(
status
)
2024