Hello; In general, when a company writes a game (MMORPG) that has zones like Everquest, DAOC, etc. Do they model the zones in a 3D package like Max or Maya and then export the vertex, texture, etc. data to a custom format for use in the engine? Kind of wondering as after I get out of college I hope to find a job working on a game of this type. Modeling or engine design preferred,and I am currently trying to learn how they do this. I''ve got a great understanding of Max and know C++, some GL and some Direct X as well as a few other languages, and now I am trying to apply some of the knowledge. Just curious if these guys write some sort or viewing software that allows them to import a 3D model and run around it as in the game to make sure all of the elements like houses, trees and grass are all in the correct place. Just the world, no NPCs or even clickable objects like doors for opening. Probably just a camera that stays a fixed distance above the ground and has the ability to turn and look around (up/down included). I''d like to play with this an see how it''s done. I can make the models in 3D Studio but I am wondering if that is the best approach. I think what mostly confuses me is, if I model a zone (or world) in 3D studio then it seems as if the engine would have to load all of the data for that zone at once. Playing DAOC it is apparent that they are using some sort of adapive mesh that adds detail as the landscape draws closer and that the mesh only loads as it is needed unlike Everquest which apparently loads all of the data for an entire zone at once. It doesn''t seem as though modeling a world in MAX and then using that mesh data in a game like DAOC would allow for this ability whereas a system like Everquest (without terrain tesselation(correct term?)), would benefit from this type of model. So, any ideas how the big guys do it? Some place to start so that I can learn how they actually design and model their worlds and if they use some algorithm on a Max or Maya world mesh that adds and subtracts detail as necessary? Thanks, Will

Probably each one of those games has it''s own "editor" for loading files, placing them, and making sure everything is in in the right spot. They also usually are used to place special zones, triggers for the AI to use, and to work with materials.

As for the details, I assume that they use some sort of algorithm to adjust the model, but don''t hold me to that.

Also, models would be created seperately in Max, and then duplicated in the editor. So instead of loading terrain with 50 trees, you load the terrain, load 1 tree and just draw it 50 times.

------------
MSN: nmaster42@hotmail.com, AIM: LockePick42, ICQ: 74128155
"It''s all part of the conspiracy of conspirators conspiring to conspire their own conspiracies..."

So that sort of editor would be used in a situation where you have a terrain with overhangs and such as you could physically model the ground with hills, valleys and cliffs and then load the ground into the editor and place individual models of trees, etc as well as textures. I hadn''t thought about things like AI paths and other special areas an still don''t see how this would work as that seems like something that would have to be specifically programmed. Guess I''m missing something there.

Are there any packages on the market for things like this? I''m assuming EQ and DAOC each use their own custom level editors and a company writing a game would likely have to write an editor that suited their needs. Is this a correct assumption? Almost seems as though they''d be writing the graphics engine before actually writing the engine.

I mean, If I write a level editor/viewer that allows me to place the groundplane and then populate it with buildings and trees then be able to roam the level as in a real game without any NPCS and particle effects then it seems as though that would include the colision detection as well as camera/movement dynamics as they would be implemented in the game. Maybe I''m just shooting in the dark.

Kind of thinking of taking a stab at a level editor/viewer to see how it would work. Any good tutorials/books on this subject available?

Thanks,

Will

I suspect some level-viewers would skip gravity and possibly even collision detection simply so that you can get a feel for the level without having to carry out pixel-perfect jumps in order to access secret/hidden areas... A similar argument applies to areas designed to be reached by rocket-jumping or by collaborative gymnastics - get half a dozen characters into a "human" pyramid...

Of course, if the entire level is designed to be easily accessible by a relatively unskilled player, then there''s no reason not to include gravity and collision detection modules of your game engine...

Eventually, you''re going to want to test your levels for real anyway, so it might be worth considering a minimal viewer to start with, with modular improvements.

So basically, if done correctly, thelevel viewer can actually be a large part of the actual engine. First add functionality to roam around unlimited and simply view the level with all the physical amenities and then start adding functionality for particles, collision, gravity, and other things so desired in a game.

Sounds like a good place to start fiddling.

Thanks,

Will