Take a look at any modern game engine and you will soon arrive at the conclusion that they look more like a black box or a set of components that cooperate to perform tasks like rendering, physics, textures, bones, etc and also provide some modding capabilities.

Unfortunatly, in the majority of cases one or more of these features will not be as good as “in that other game engine” which is a result of either bad implementation or simply trade-off between requirements. Additionaly, most of those engines are difficult and messy to extend in order to accommodate future changes in hardware or software. Ultimately, those engines have to be rewritten from scratch in the worst case .

For a game developer, finding a game engine suiting the specifications of the next project is a balancing act and a decision to be taken as early as possible in the project's lifecycle making this one more thing to worry about instead of focusing on the game's content. The most likely case is to have come up with 2 or 3 choices and wishing to have the option to mix and match the best components from each engine. This is clearly a design problem caused by the way game engines are designed today. To make matters worse, software companies sell licenses at high prices which makes it difficult for smaller groups of developers / software companies or even creative individuals to get started buidling their project. Creative groups or individuals might consider creating their own game engine which means spending a considerable ammount of time in designing and debugging a rather complex piece of software.

These design choices also have an impact to the game player / power user / creative individual who will also have to deal with the benefits and disadvantages of different game engines used by the games they have installed on their machine. For example, it so happens sometimes to have two games based on the same game engine not neccessarily both running on a specific machine. Choosing a game based on a closed design means locking down to the set of customisation options it has to offer.

Closed designs are not the only option of course. Because of the collective nature of open source projects, open source game engines tend to be much more modular but still, most choices can only be applied during compilation time. However, people have being programming like this for ages, producing better software than a closed design but still not good or flexible enough for taking game design one step forward.

Drag[en]gine is a free software project with a highly modular structure. Its design is based on the idea of treating it as if it was an operating system.

The entire game engine functionality is packaged into Modules which have a similar role to that of device drivers in an operating system. The engine itself is the analogue of the system kernel managing the modules, resources and providing a straightforward abstraction layer to the underneath operating system.

Modules provide all the features other game engines hide as built-in features ( or features they lack ). Due to the loose coupling of the modules with the system and other modules it is very easy to exchange or improve a module without interfering with the rest of the engine. As a result the modularity extends from the developer to the end user who can now choose the optimal module combination for his personal computer even down to per game setups to maximise performance and user experience.

Developers do not have to worry anymore about what graphic routines or physics library animates the scene's objects which helps keeping them concentrated on the content and special functionalities of their game.

User customisations can be performed even at run-time deciding what graphic renderer or physics library works best on their own machines!