TripCom and Triple Space Computing: Web Services Built on Persistent Publication

What was TripCom?

TripCom, short for Triple Space Communication, was a European research project on a different way for software to exchange data. It received funding from the ICT strand of the European Union’s Sixth Framework Programme. It finished in 2009. Its subject was “triple space computing”. Applications would write and read machine-readable data in a shared, persistent space instead of sending messages to each other.

What problem did triple space computing address?

The project started from a comparison with the Web. A web page is published once and persists. Anyone who knows its address can read it later, and the publisher and reader never need to be online together. Many Web services built on WSDL and SOAP worked differently. They needed sender and receiver to be connected at the same moment, to know each other, and to agree on a data format in advance. Add-ons such as WS-Addressing, WS-Notification and WS-Eventing, or SMTP in place of HTTP, tried to add asynchronous delivery. The project argued for a more fundamental change: apply the Web’s “publish persistently and read” model to machine-to-machine communication.

How did it work?

The idea combined tuple spaces, shared object spaces, persistent message-based architectures and RDF. In tuple space computing, processes communicate by reading and writing data structures in a shared space, not by exchanging messages. Triple space computing made the stored unit an RDF triple. RDF represents data in subject-property-object statements. The shared space therefore held statements whose meaning machines could interpret.

The project described three kinds of decoupling:

• Reference: communicating processes did not need to know about each other.
• Time: processes did not need to run at the same time.
• Space: processes could run in different environments, as long as both could reach the same space.

The project also aimed at Internet scale. It planned to build on HTTP, RDF and commercial storage such as databases and file systems. These are the same ingredients used for web pages.

What was built and tested?

The project produced a kernel, the core software of a triple space. A research paper describes it as having a query processor, a security manager, a transaction manager and a web service API. The same paper notes that its modules could run on different machines and that it targeted powerful hardware.

The use cases were varied. One case study outlined a European e-health information system. Medical professionals would retrieve records held in distributed hospital databases. Other uses included enterprise application integration, virtual marketplaces for e-commerce and mobile computing services. The project also published ontologies for business data exchange standards. One aim was to define ontologies for EDI, covering syntax and semantics, to overcome heterogeneity.

Where do the ideas appear afterward?

The project partners collaborated in a follow-up project, SOA4All, which aimed to use triple space technology to connect very large numbers of distributed web services. Later academic work on semantic middleware for the Internet of Things also cites triple space computing, and names TripCom as one of two main implementations. That work notes the kernel’s server-centered design as a limitation for mobile and embedded devices.