Most online services begin by creating another account and storing another copy of our data. The service controls the identity, the data, and the rules for accessing it.
Keeping data on a personal device or private server gives us more control, but it creates another problem: how can people, devices, services, and agents use that data across different networks without exposing the entire device or private network?
DHttp is an open-source infrastructure stack built around this problem.
In DHttp, an endpoint can have its own stable name and cryptographic identity. The identity is controlled by its owner, and the private key stays on the owner's device. A computer, device, service, or agent can be recognized across different networks without creating a new identity for every platform.
Being reachable does not mean being public. APIs are private by default, and the owner decides which capabilities another identity can use.
For example, a family member could control the lights and thermostat in a guest room without gaining access to cameras, door locks, personal files, or the rest of the private network.
This is different from a VPN. A VPN places devices inside the same private network. DHttp keeps endpoints independent and uses identity and API permissions to decide who can do what.
Developers can build extensions on top of this framework. An extension runs in the user's environment, uses the existing identity and permission system, and keeps its data there. Developers can focus on the service they want to create instead of rebuilding accounts, connectivity, and access control each time.
The goal is an open ecosystem where devices, services, personal data, and agents can work together without giving control to a central platform.
What would you build on this model? Which part would you design differently?
Top comments (0)