Is it really that hard to understand the difference between the drawing distance and dynamic removal of objects based on priorities? No, it doesn't depend on the definitions of 'remove' and 'sub-objects'.
Obviously you are a not a programmer; Yes the definition of "remove" and "sub-objects" is important. Are objects being removed from memory, from video memory, simply being turned invisible or something else entirely? Are the objects just multiple models or are they something else.
Oblivion for example displayed objects based on the user's setting: only show elements in group X once camera is within range, only show elements in group Y once camera is in range, only show elements in group Z once camera is in range. Sim City 4 as well did this: cars and pedestrians (IE: objects that add more detail) only began being displayed once the camera was within a certain range.
At any rate you completely failed to offer a single constructive point to a conversation that had nothing to do with you. Keep your mouth closed next time (or more accurately, keep your fingers motionless).
Actually, they load different 'versions', for lack of a better word, of an object depending on your distance from said object, thus increasing detail surrounding your place in the environment while decreasing detail that you're most likely unable to notice at a distance. The same thing is done with textures. This is where 'pop in' comes from. It's not modifying an object on the fly to meet the needs of the system.
What this engine does is based on your system's performance it determines what objects to render and what to ignore entirely as well as their quality. In GTAIV, for example, you'll be able to set the distance that the object is rendered and the quality (I haven't played GTAIV on the PC myself) of the texture rendering and perhaps to geometric quality. To achieve something similar to the Game Engine Elemental uses, you'd need to be able to determine what objects are rendered and their specific details. For example, in GTAIV, how detailed are the car models? How detailed are the headlights, wheels, rims, seats, radio, etc.? How many pedestrian configurations are used? How many will be holding objects like drinks or cell phones, and how many different objects will be used in the object pool, how complex will they be rendered, how complex will the pedestrians shoes, hats, gloves and jackets be rendered? Will it use its resource draining traffic pattern system, or a more simpler and less realistic randomised system? This allows the system to scale to a much more specific degree than the pathetically simple "Very Low, Low, Medium, High, Very High" settings we're given these days. The benefit of this is to ensure as many people as possible can play the game with the details that they considered to be important set to their highest possible setting with the details that they consider to be less important rendered at a lower quality.
I'm familiar with how the other mentioned engines work (in regards to how they scale quality / detail) to a certain degree - and you're right. It just seems that Elemental's engine is simply taking away the settings (from games such as GTA IV) and is instead setting everything relating to draw-distances internally within the actual engine. The engine could be taking a survey of video memory and scaling the settings based on that alone. This is why I was asking what the difference was - not to be rude or because I was unfamiliar with how game engines work, but because I was curious.
