Actaully, this does sound intersting. Josh, are all your "OpenGL" based DLL's made in pure basic? If so, are you interested in having any/all of them converted to C/C++?
As for open source, will it go by GPL or LGPL? I like the idea of if any changes are made to the library itself then you must share the changes of the libray itself, not your whole program idea.
As for sound, check out OpenAL. It does a good job as far as I can tell from how the americasarmy.com FPS game uses it. It might be worthy of checking out.
As for as my library I'm doing, it's basically being programmed with the idea of becoming a "Nuclear Glory" game library for Blitz+ having some blitz like "entity" qualities. Anyways, I'm confident of working with OpenGL now even more so than with all the years I've toyed around with DirectX. Also, being able to work with C/C++ dlls, we dont really need any use of Banks as in to store graphics, 3d models, sounds, music, etc... With C/C++, vector arrays are a god send for easy control over various data and memory.
About the collision, I was going to approach it first by doing sphere-to-poly collision and then add in ellipsoid type method simply by multiplying or dividing the polys that are tested against the sphere. The group of polys that are to be tested would be scaled the same for the x,z plane (ie: ellipsiod width) and scaled along the y plane (ie: ellipsiod height) to give an ellipsiod effect. Ofcoarse, the sphere would be positioned correctly via the ellipsiod width and height scales when testing collision against polys. Basically, the world changes, not the sphere.
Or, I go about the fluid studios way or the varients derieved from it for collision. I'm just spouting ideas here. Doesn't mean I'm "wanting" to do it for this project as you seem to have interest in figuring it out for this project. Ofcoarse, I don't mind helping out with the idealogy behind it and/or doing it myself.
I'm interested in helping out somehow. I can get the dll code in C/C++ form as the Visual Studio 6 Enterprise compiler can optimize it for speed. I can also showcase what I have with my library if there is anything in there that might be of interest to include in this project. I really do want to see this open source project happen. Are you going to have a website for it or keep details here on the forums? I just don't know where to start or what to do to add to the project. I'll get what I have up asap at least for consideration to include anything from it.
Oh, also I use the devIL library for including images/textures. The idea behind that is to make it easier for me to include "Load Media From Memory" or "Load Media From Pack Files DIRECTLY" type of functions when I get around to it. That idea is something I would like to see this open source project include off the bat. Blitz seriously suffers from this directly.
If any ideas I have above sounds useful in anyway, I'll be glad to help out. I just would like to see more in detail whats needs to be done, what's are already done, who's doing what, all that good orginization stuff... before I seriously dive into it.
As for open source, will it go by GPL or LGPL? I like the idea of if any changes are made to the library itself then you must share the changes of the libray itself, not your whole program idea.
As for sound, check out OpenAL. It does a good job as far as I can tell from how the americasarmy.com FPS game uses it. It might be worthy of checking out.
As for as my library I'm doing, it's basically being programmed with the idea of becoming a "Nuclear Glory" game library for Blitz+ having some blitz like "entity" qualities. Anyways, I'm confident of working with OpenGL now even more so than with all the years I've toyed around with DirectX. Also, being able to work with C/C++ dlls, we dont really need any use of Banks as in to store graphics, 3d models, sounds, music, etc... With C/C++, vector arrays are a god send for easy control over various data and memory.
About the collision, I was going to approach it first by doing sphere-to-poly collision and then add in ellipsoid type method simply by multiplying or dividing the polys that are tested against the sphere. The group of polys that are to be tested would be scaled the same for the x,z plane (ie: ellipsiod width) and scaled along the y plane (ie: ellipsiod height) to give an ellipsiod effect. Ofcoarse, the sphere would be positioned correctly via the ellipsiod width and height scales when testing collision against polys. Basically, the world changes, not the sphere.
Or, I go about the fluid studios way or the varients derieved from it for collision. I'm just spouting ideas here. Doesn't mean I'm "wanting" to do it for this project as you seem to have interest in figuring it out for this project. Ofcoarse, I don't mind helping out with the idealogy behind it and/or doing it myself.
I'm interested in helping out somehow. I can get the dll code in C/C++ form as the Visual Studio 6 Enterprise compiler can optimize it for speed. I can also showcase what I have with my library if there is anything in there that might be of interest to include in this project. I really do want to see this open source project happen. Are you going to have a website for it or keep details here on the forums? I just don't know where to start or what to do to add to the project. I'll get what I have up asap at least for consideration to include anything from it.
Oh, also I use the devIL library for including images/textures. The idea behind that is to make it easier for me to include "Load Media From Memory" or "Load Media From Pack Files DIRECTLY" type of functions when I get around to it. That idea is something I would like to see this open source project include off the bat. Blitz seriously suffers from this directly.
If any ideas I have above sounds useful in anyway, I'll be glad to help out. I just would like to see more in detail whats needs to be done, what's are already done, who's doing what, all that good orginization stuff... before I seriously dive into it.