• Advertisement

rgirard413

Member
  • Content count

    112
  • Joined

  • Last visited

Community Reputation

133 Neutral

About rgirard413

  • Rank
    Member
  1. Collision detection

    http://www.peroxide.dk/download/tutorials/tut10/pxdtut10.html :)
  2. don't forget in some cases you wouldn't even need a flag, if it has to do with turning an item in to an npc to proceed. the only way that npc will talk to you is if he gets that item....
  3. Use HLSL?

    So is designing a new project with dx9 even worth it at this point? Or will dx10 support applications written in dx9?
  4. Use HLSL?

    I was curious from a game design standpoint, is it highly beneficial to develop a game, or even a demo with HLSL being used, or do most of the popular games(like WoW for example) just run off normal hardware techniques? Basically, I know I'm not going to create a popular game or anything like that, however I would like to create some kind of demo or mini game for experience, but would perfer to use the same technologies so I can stay current.
  5. this isnt an answer to your question but i have to ask. Why are you looping through all the triangles twice? I would think you would just loop through them once and if they are in your bounding box, put them in this leaf and increment the triangle count. then if you have a triangle count, fill your vertex buffer. Maybe im wrong though?
  6. doesn't this work if you don't want to use vector...? type **map; map = new type[number_x]; for(int i = 0; i < 10; i++) map[i] = new type[number_y]; yes its ugly and why wouldn't you want to use vector...
  7. dynamic object usage?

    is there a way in to addition of doing this that i can access the functions of the driver individually without using virtual functions. So say i declare p as new a(); std::auto_ptr< Parent > p = Parent::create(which); p->x(); but then i can do something like: p->CreateVertexBuffer(params...); if i am using the dx driver or p->whateveropengl'sfunctionis(params...); Essentially I wont want to have to declare virtuals for EVERY function...
  8. dynamic object usage?

    thanks a lot guys, this was actually my way of creating a dynamic allocating graphics driver(dx vs gl) so that the user can specify which they want to use and i wont have to change the rest of my code :)
  9. dynamic object usage?

    is there a way i can pass say a value to p to have it determine which to use like: Parent *p = new Parent(which); if which is 0 it will automatically use A, otherwise B?
  10. I couldn't seem to figure out how to do this. basically i have two classes(a and b, both of which share same function names, but may have different variables.). what im trying to do is, based on a separate value, call one or the other classes. ie. Class A { public: int x(); private: float somevalue; }; Class B { public: int x(); private: int someothervalue; } main() { psuedo - enter some value, if its a 1, use class A, if its a B, use class b //(class) has to be dynamic as to avoid using switches or if statements // all the time. regardless of hte value i enter, ill always use // (class) as the object. (class).x(); } Is this something that would be achived through inheritence? Any help would be great.
  11. I'm intersted in method others have used to define the relational method between the collision system and the terrain data. Logically from my point of view, I have had a terrain class that loads and stores some data and a function to return the terrain data(just the x, y, z's of the triangles verts. Then with the collision system, I pass to it the data returned from the terrain function and process it in the separate collision class. I'm curious how other approach this process and looking for a more efficient method or input on my current approach. Thanks!
  12. Basically I'm just looking for clarification that my process/ideas for implementing this are right. I'm going to be applying this to a heightmap for the time being: Should my quadtree class be the class that actually loads the heightmap? The heightmap is blocked off into section(say heightmap is 128x128 and each section is 16x16, which would give me 8x8 sections. each of these sections are which actually hold the triangle data? During heightmap loading, I should apply quadtree conditions/recursive traversal to it, such that the appropriate triangles for the sections are added to the correct section? During rendering, I traverse the nodes, ensure that each node is in camera view, if it is, advance to child nodes and repeat. Once we are at the base nodes(the smallest that need to be drawn based on some frustrum conditions), we render all subsequent triangles in that section, and that section is what is associated to that leaf of the node? I think this is everything, I just want to make sure i have a clear understanding of what is going on and how to interprete the data properly. Thanks for all your help.
  13. I remember having the same problem, unfortunately I'm at work right now so I wont be able to view my code to see what I did(was quite a while ago). When i find it, ill post help if no one else does in the mean time.
  14. who wins, who loses, you do the math!

    i think you need more variables to determine the outcome, for example "track complexity" or something. Is this a dragway they are racing on, in which having more power will more than likely be a huge determining factor. Or is it a very winding track with many curves, where handling and driver skill will be more beneficial. But of course two cars with similiar power is going to more than likely come down to driver skill on a straight track...
  15. QuadTree Implementation?

    Benny: That helps a lot. Is this where the bounding boxes come into play, if all three points of a triangle are inside bounding box of this this node? where does the leaf come into play?
  • Advertisement