Graph languages are a fascinating new technology proposed by Project Samarai. At the core, graph languages are information modelling languages, powerful tools to express, store and process complex digital information. But they are also programming languages, allowing end-users to create new functionality. Together with versatile application frameworks and runtime environments, they become highly flexible communication and collaboration platforms. Integrated modelling environments enable CAD for arbitrary information. Graph languages could become key-technologies in future artificial intelligence systems.

Graph languages are unlike any other language. Instead of characters, they use semanto-relational information bits; instead of words, they use reoccurring patterns in semantic graphs. No human will ever speak a graph language directly. Yet, they will enable computers to understand us much better. Ultimately, they have the potential to fundamentally change how we digitally communicate and collaborate with each other. They will allow us to use the full potential of computers.

Havel, Project Samarai’s primary graph language, is a versatile, flexible and modular language inspired by fundamental human logic. It allows for almost unlimited information complexity, multi-dimensional information, and automatic information trustworthiness assessment; encryption and multi-level data access security are integral parts. Maybe most importantly, Havel is not only an information modelling language but also a programming language. A programming language that seamlessly integrates with digital information, enabling radically new approaches to application design: Information-centric computing. Together with its runtime environment, Havel enables protocol-free communication, collaborative computers and semantic social networks.

Havel stores information using SRDM, the semanto-relational data model. SRDM can be described as an extended, resource and performance-optimized version of RDF, the underlying technology of the semantic web. SRDM is fully semantic and has highly desirable properties, i.e. structureless/chaotic repositories, automatic information merging and self-descriptiveness. Because of its application and schema-independence, data expressed with SRDM is called universal data.

Havel is owned and coordinated by Project Samarai, a non-profit organization based in Switzerland. It is, together with all its sub-projects, open source and the code will be published as such in due time.