Hello everyone, Just wanted to get your response on the followung idea of a middleware product, which has some proof-of-concept implementation behind it. The main idea is that the middleware I am talking about implements 'loadless' landscape LOD visualisation. That means it is possible to place the camera on the orbit of a planet, and then drive it down to surface and up again and without any nagging "Loading, please wait..." screen. The implementation does not need 100's of megabytes of pre-calculated data, it is calculated in runtime instead. The following features are also planned to be developed: 1) Support for space bodies like star, planet, satellite, asteroid, comet, etc. (with irregular shapes, too) 2) Metric/real world sizes, game time 3) Physical simulation of star systems 4) Detail surfaces, dynamic per-pixel lighting model & shadows + self shadows, horizon mapping 5) Time of day with weather, volumetric clouds, etc. 6) Water surfaces including realistically flowing rivers 7) Grass & trees 8) Negative declivities, holes (aka "sci-fi landscapes") 9) Volumetric fog 10) Dynamic night sky The stuff above will be accompanied by editors and pseudo-random landscape/texture generator. The mega-goal is to implement the above features as middleware compatible with major engines. In a more distant future a custom SM 4.0 based reneder will probably be done. Could you guys comment on the idea? Is this sort of thing needed for game projects and is there a market niche for it? Why? -- regards, serge

Hi Serge, you asked:

>Is this sort of thing needed for game projects

The fact that you have to ask this question should tell you that you're probably approaching this the wrong way. You're starting with an invention before determining if there's even a need for it. Best to start by knowing about a need, and then come up with a way to satisfy that need. Are you building this tool (sorry if that's not the right word to characterize it), or contemplating building it, because you were developing a game and needed this set of features? If so, and if you would rely heavily on your own middleware to build a number of other games too, then you'd know you're on to something. In fact, you ought to be your own first customer. After building your middleware, you ought to build some games with it - that'd be the absolute best way to ensure that it's useful for others as well.

To create a game that will need all the features I quoted it is way too much for poor me :-) Also, I cannot cover all targets that this middleware can power - (mmo)rpg, space/flight-sim, rally, etc. That's why I approach it this way.

So I am asking if somebody with relevant experience can share his/her thoughts with me whether this sort of thing can save time and money for the game developers.