|
Panda3D is a game engine that includes graphics, audio, I/O, collision detection, and other abilities relevant to the creation of 3D games. Panda3D is open source and is, as of May 28, 2008, free software under the revised BSD license. Releases prior to that date are not considered free software due to certain errors in the design of the old Panda3D license. Despite this, those older releases of Panda3D can also be used for both free and commercial game development at no financial cost. Panda3D's intended game-development language is Python. The engine itself is written in C++, and utilizes an automatic wrapper-generator to expose the complete functionality of the engine in a Python interface. This approach gives a developer the advantages of Python development, such as rapid development and advanced memory management, but keeps the performance of a compiled language in the engine core. For instance, the engine is integrated with Python's garbage collector, and engine structures are automatically managed. The manual and the sample programs use Python, although the developers are working on translating the manual to C++ and providing C++ sample programs. A developer using Panda3D typically writes code in Python, but it is also possible to directly access the engine using C++ code. The users of Panda3D include the developers of several large commercial games, a few open source projects, and a number of university courses that leverage Panda3D's short learning curve. The community is small but active, and questions on (the forum ) are generally answered quickly. ==Design== Panda3D is a scene graph engine. This means that the virtual world is initially an empty Cartesian space into which the game programmer inserts 3D models. Panda3D does not distinguish between "large" 3D models, such as the model of an entire dungeon or island, and "small" 3D models, such as a model of a table or a sword. Both large and small models are created using a standard modeling program such as Blender, 3ds Max, or Maya. The models are then loaded into Panda3D and inserted into the Cartesian space. The Panda3D scene graph exposes the functionality of OpenGL and DirectX in a fairly literal form. For instance, OpenGL and DirectX both have fog capabilities. To enable fog in Panda3D, one simply stores the fog parameters on a node in the scene graph. The fog parameters exactly match the parameters of the equivalent calls in the underlying APIs. In this way, Panda3D can be seen as a thin wrapper around the lower-level APIs. Where it differs from them is that it stores the scene, whereas OpenGL and DirectX do not. Of course, it also provides higher-level operators, such as loading models, executing animations, detecting collisions, and the like. Panda3D was first engineered before the existence of vertex and pixel shaders. It acquired support for manually written shaders in 2005. However, users have been slow to leverage modern per-pixel lighting techniques in their games. The developers theorize that this is because shader programming can be quite difficult, and that many game developers want the engine to handle it automatically. To remedy this situation, the Panda3D developers have recently given Panda3D the ability to synthesize shaders automatically. This synthesis occurs if the 3D modeler marks a model for per-pixel lighting, or if the modeler applies a normal map, gloss map, self-illumination map, or other capability that exceeds the capabilities of the fixed-function pipeline. The intent of the synthesis is to render the model as the modeler intended, without any intervention from the programmer. 抄文引用元・出典: フリー百科事典『 ウィキペディア（Wikipedia）』 ■ウィキペディアで「Panda3D」の詳細全文を読む スポンサード リンク
|