The "Glue-ML language" is a set of schemas for describing systems and connections.
In particular it is used to describe:
I) Assembly and configuration of data pipeline structures
II) Communicating state machines, with transition guards and transition actions
As of 2016, the principal schemas of the Glue-ML language are defined using the following OWL (-Lite/RDFS) ontologies:
These schemas are concerned primarily with category I features, i.e. defining pipelines. Meanshile category II (state machine) structures are more purpose specific, and thus are built as activity sites within such pipelines.
Other ontologies are also used to specify schemas for UI features and particular application scenarios, which is where explicit state machines come more into play.
We call Glue-ML a (pseudo-) "language" because it defines a programming model associated to a vocabulary. If Glue-ML were to be written out as "code", the syntax of Glue-ML would be any syntax compatible with RDF. But more commonly, one works with a set of graphs (in files or databases) which together specify the required computation and application features for your application. These graphs may be edited with a wide variety of tools.