From: VS R. <aki...@ya...> - 2001-07-31 06:13:41
|
Norman, I opted for global coords initially so that I dont have to delete the = node from the tile each iteration and then create the node afresh with = updated tile coords. This in my opinion would be more expensive = computationally especially when we think of using more moving objects in = the scene (say 32 permitted by the ssgSelector node). As far as accuracy = is concerned I dont think there has been any compromise since all the = transformations contained in the worldcoordinate function are used to = arrive at the transformation matrix of the object. As far as HOT testing is concerned I see that presently only world = position and world transform of the aircraft are available in viewer. I = intend searching plib to find the easiest way to compute the appropriate = viewing matrix, viewpoint and direction for use with the carrier in = global coords. If you have the time I could do with some ideas. Regards Ranga ----- Original Message -----=20 From: Norman Vine=20 FWIW Making another SSG scene is pretty easy and walking it will be real fast as there will not be much in it. =20 I would do this using the FGFS tile based coordinate system=20 =20 The trickiest thing I think will be keeping track of what tile the Object is in as it moves. IMHO you do want to use the FGFS tile based coordinate system as it is MUCH more accurate then trying to use a global coordinate system and using it will allow you to use the existing per frame=20 'tile transformation setup code' and hit_list code almost as is.:-) =20 |