At this point, we offer a definition of Domain-Specific Language:
- A Domain-Specific Language is a custom language that targets a small problem domain, which it describes and validates in terms native to the domain.
Most computer languages are textual, with their statements and expressions consisting of sequences of characters in a standard character set. Graphical languages have become increasingly popular in recent years, particularly with the emergence of the Unified Modeling Language (UML) as a popular set of standard notational conventions for depicting elements in an object-oriented software system.
When computer experts talk about languages, they usually mean general-purpose textual programming languages such as Visual Basic, C#, or Java. In Domain-Specific Development, our interpretation of the word language is widened considerably—it includes graphical languages such as UML, flowcharts, entity-relationship diagrams, state diagrams, Venn diagrams, and so on. We also include other textual languages such as XML and domain-specific varieties like SQL and regular expressions. We even think of tabular and form-based formats such as spreadsheets or the Windows Forms Designer as being languages. Special languages also exist for domains such as music notation and direct-manipulation interfaces. With the power available in modern computers, there is absolutely no need to be restricted to simple linear textual notations to convey our intentions to the computer; we want to exploit the power of the computer to provide means to express the author's intent as directly as possible, thus increasing the efficiency of our development. This includes interactive aspects such as dragging and other gestures, context menus, toolbars, and so on.
There are two main forces at work driving the evolution of languages. The first of these is the progressive lifting of the level of abstraction at which we express our intentions about what we want the computer to do. Originally, programmers had to express their algorithms and data structures in terms directly accessible to the computer hardware, which was efficient for the hardware but very tedious and error-prone for the programmer. Subsequent developments such as symbolic assemblers, filing systems, third- and fourth-generation languages, databases, class libraries, and model-driven development have moved the languages in which developers express their intentions further from the computer hardware and closer to the problems they are trying to solve.
The second force driving language evolution is the increasing variety of available digital media. Originally, computers were used purely to compute with numbers, then also with symbols and texts, and then with bitmaps and images. The evolution of computing has reached a point where the limitation on how we express our intentions is no longer the physical capabilities of the computer itself but the limits of our understanding of how to construct and manipulate computer languages. In Domain-Specific Development, instead of building on a general-purpose language in order to solve a problem, we use a language that is itself designed to suit the problem being solved.