Skip to content
REAL TIME WEB

UNDER THE HOOD

How it works

Live data channels, user-controlled sources, and coexistence with the World Wide Web.

A framework for the Internet of Everything

The Real-Time Web is intended to provide a more flexible way to build Internet of Everything systems.

Today, sensor data is commonly collected through proprietary APIs and middleware before being copied into databases or cloud platforms. Applications must then request the stored data through platform-specific interfaces. This creates additional infrastructure, repeated integrations, duplicated data, and dependency on intermediaries.

Direct live channels

RTW proposes an alternative. Live data channels can connect the original data source directly to the appropriate web service, application, AI agent, device, or web page for a particular user instance.

Figure 3-1 concept: Bijective Network System (redrawn from source description)
World Wide Webdocuments · mediaReal-Time Weblive bidirectional channelsPhysical worlddevices · sensors · peopleDynamic data channels connect IoE resources to a service, application, or pagefor a specific user instance, concurrently with the existing WebInternet of Everything participants

Figure 3-1, as described in the source, shows the Real-Time Web connecting the digital Web to the physical world through a Bijective Network System. Dynamic data channels connect Internet of Everything resources directly to the appropriate service, application, or web page for a specific user instance. The resulting network operates concurrently with the World Wide Web, adding live, bidirectional communication.

The Real-Time Web does not need to replace the World Wide Web. It can operate alongside it in symbiosis. The conventional Web remains well suited to documents, published information, and media, while RTW adds a dynamic network for live resources, machine interaction, and continuously changing data.

Bidirectional addressing

Each participating endpoint has a unique corresponding representation in the network. Authorized updates can be distributed as they occur, rather than being retrieved only through a request.

Data can remain associated with its original source. Network relationships can persist across participating services. Data access and transformation logic can be controlled separately.

Four layers in every exchange

Even a simple action like sending a message involves all four MAD layers, often with different owners. See Morphic Architecture.

Infrastructure

Hardware / connectivity provider

Operate devices and transport the message

Network / security

Network operator

Identify, authenticate, and protect the communication

Domain / application

Application developer

Construct, process, send, and receive the message

Resource / ownership

End user

Create the message and decide to share it

Shared state

MAD introduces Ghost Space: a shared-memory representation of a network resource through which infrastructure, security, applications, and owners can cooperate without one party owning the entire stack. See Ghost Space.
Endpoint ↔ Ghost: unique bijective relationship
Ghost Space: shared coordination environmentEndpointsensor · device · agent1 : 1addressable from both sidesGhostshared-memory representation

How participation works

Understanding the architecture is not the same as running a live service. Ownership, identity, and access still follow MAD: a domain is an address space, a ghost stands in for one endpoint, and morphing grants read access that can be withdrawn.

MAD
Want an introduction to Morphic Architecture Design?
Read Morphic Architecture →