While I'm not sure there is a distinction any more, I think it's important to point out that the design goal of OpenDX was to provide researchers with a functional, highly interactive, flexible tool for visualization.
To produce the computer graphics special effects we see in the movies and on TV, three essential tools are used by animators: modeling tools, animation tools, and rendering tools. Modeling tools are used to construct objects (chiefly, dinosaurs and space ships). Animation tools permit the animator to make the models or the camera viewing the models to move about in the scene. Renderers, like photographic cameras, create 2D images of the model world as seen through the perspective of the virtual camera.
In comparison to the output of "entertainment" animation packages, far less attention was paid to the final output quality of the visual imagery achievable with OpenDX. This is not a slur against all the hard work done by the developers, but frankly, OpenDX's imagery does not stack up against the elegant high-quality, ray-traced, reflective, antialiased, texture-mapped images we see in big Hollywood productions. On the other hand, do you have any idea how long it takes to render images like that? (The answer depends on the complexity of the objects in the scene and the size of the final image, but suffice it to say, it's not interactive!)
Furthermore, while OpenDX provides a simple animation capability which is more than sufficient for the majority of users, anyone who has ever used a "real" animation program will be frustrated by the lack of tools for producing more complex animation. If the word "keyframes" means anything to you, OpenDX does not have them; if keyframes means nothing to you, you don't need them.
Finally, OpenDX provides virtually no facility for creating models by hand: it is neither a CAD (computer-aided design) program nor a dinosaur-drawing program. It's possible to import simple models from other programs, though currently, there are few filters available to convert the output of mainstream modeling programs into DX format. However, the extensibility of OpenDX and the availability of its data model and object description certainly permit anyone so motivated to write such an import program. These may sound like strikes against using OpenDX, but in fact, we've rarely needed the above mentioned features (or maybe we've just learned to live without them).
Again, though, I emphasize that the target users of OpenDX are researchers who want to visualize data-driven (not hand-drawn) objects. There is the potential to be sucked into the "glitzy" end of presentation when high-powered animation tools are available, such as those found in products like Maya, Softimage, Lightwave, and 3D Studio Max. Look at the bright side: OpenDX saves you from the temptation of getting too glitzy! The output is more than adequate for most people and later in this workshop, I'll mention some tricks you can use to improve on the standard quality output that you get from OpenDX. But these remarks should not dissuade you from continuing: there are few other "scientific" graphics packages that compete head-on with the "entertainment" packages either. One could argue that if the imagery looks too good, it loses credibility precisely because it looks like some "artiste" got hold of it and he or she may have "interpreted" the underlying science it is supposed to depict. Quite frankly, DX images look like they are made by scientists. Homely but credible.
