The control-command environment of the Virgo experiment

The Virgo experiment is a challenge in many aspects, particularly for the control-command environment. The essential requirements for flexibility and reliability, induced by the very drastic technical constraints as well as the unapprehended nature of the overall logic imply that strong and well-defined paradigms are used throughout the building phases of this environment. This results in providing a software architecture, called the Supervisor environment based upon some basic principles such as the component oriented packaging or the use of scriptable formal languages. In addition, keeping this architecture simple and well defined makes it compliant with quality requirements for long term maintenance, evolution and interoperability. This environment provides the basic services required by user applications to operate the experiment such as communication, database management, error handling, logic and graphical control as a set of interoperable components making extensive use of the object oriented paradigm and of the scriptable specification languages. The Supervisor itself is the basic control application and can be cloned, refined and customized to extend the basic logic towards specific needs. The paper shows some of the basic ingredients of this architecture, and discusses how user control applications have been integrated in this environment.