Quote: "Think outside the box. All segments are is geometry put together for you already. They consist of all entitys linked together by the FPS file"
right, i also think that using entities instead of segments but with the same number of polys would not change anything, it gets even more bad because of physics-calc when you use nonstatics.
i dont think you can build a complete destroyable level because of the physics, would result in imense calculation and bad framerate.
but think this:
example: simple builded house wich you cannot enter of 5x5x5 segments, each a simple cube with 6 polys = 600 (x 2 because of inner walls) for walls and 300 for floors/roof = 1500 polys. using a simple cube instead = 6 polys.
because of the more complex segments you normally use, this is an extreme example.
but you can save the invisible parts of walls (x,y rows and inner/outer walls) that are between one segment and the next when using 1 entity instead of x segments. not much with complex segment-meshes, but everything counts, even when you just save 5 frames with a fully filled level, then you save 5 frames = good
pro: polysave for other enviroment. good for scenery-elements in the background and big surfaces. good for destroying buildings/secret rooms and more.
con: big textures limiting the maxbuild of 1800mb even more, so you have to use it not too often on too many different things.
AND you have to place it really accurate when you wont get failures. segments fit perfectly together.
if you can use this in a big scale depends on how complex your meshes have to be.
mixing it up in a balanced way is the best you can do = the art of a good level-setup/cleaning > professional game-developing.
my thoughts.
n8
