Hey guys. Last time I posted here was a good 2 years ago and I was still trying to figure out what degree to do where. Since then I've joined the *joyfull* world of instense Computer Science studies and all related matters, and am about to finish my second year. For our third year we need to do a major, 1 year long, project and Im having some problems selecting the theme. I do know more or less what I want to do, I just dont know the specifics. What I want to do is a project related to video game programming, which involves AI or physics (or maybe even graphics programming, who knows). The problem is, they obviously wont let me have a project called "Video Game Physics Engine".. So what Im asking you guys is if you know any interesting topics withing these areas that I could work on. Some of the ideas I had were something like "Interactions and modelling of large numbers of objects from a physical point of view" (got this idea from the physics and particle engines used in SW:The Force Unleashed) or "Application of realistic fluid behaviour to video games", or even "Inteligent behaviour in video games: How AI might be used in a video game to create opponents that learn from previous experience"... Those don't really sond like scientific project names, but I think you can get where Im trying to get. P.S. Btw, any of you know of any free graphics engine I could use to power my project? I was thinking of going with OGRE as it should be able to do whatever I need and I have a book explaining bits of it, but if any of you know of any other easy to use and with a good performance please let me know. Cheers, Zepee EDIT: Just in case you didn't understand, Im aiming more towards technology and engine creation than gameplay programming.

OGRE I hear is pretty good, but I haven't used it myself so I can't say for sure. I would recommend using XNA or something, since I have used it recently and found it to be surprisingly good at what it advertises. It takes care of a lot of the lower level things, but still leaves plenty for you to code yourself, including all of the important engine parts.

Just whatever you pick, make sure it's feasible. A remotely decent physics simulator that can run in real time is going to be complicated and take a lot more time than you might expect. But, any part of a game you take and try to turn into a full-fledged simulator or something is going to be that way. Physics is at least a good choice in that it would have obvious immediate academic appeal to any approving committee, and a good bit of the stuff you'd need is easily found on the internet if you have a good math background.

As for specifics, well, that is hard to say. You could at least start with the obvious if you're going with a physics engine: basic movement (velocity, acceleration, momentum, elastic / inelastic collisions, rotation kinematics, stuff like that). If you get that far, then look into the cooler / harder stuff.

Quote: Original post by Telgin
OGRE I hear is pretty good, but I haven't used it myself so I can't say for sure. I would recommend using XNA or something, since I have used it recently and found it to be surprisingly good at what it advertises. It takes care of a lot of the lower level things, but still leaves plenty for you to code yourself, including all of the important engine parts.

Yea, I have some experience with XNA, but I want to do this in C++ so I can have a major game related project in C++ in my portfolio..

Quote: Original post by Telgin
Just whatever you pick, make sure it's feasible. A remotely decent physics simulator that can run in real time is going to be complicated and take a lot more time than you might expect. But, any part of a game you take and try to turn into a full-fledged simulator or something is going to be that way. Physics is at least a good choice in that it would have obvious immediate academic appeal to any approving committee, and a good bit of the stuff you'd need is easily found on the internet if you have a good math background.

Well, whatever I pick Im supposed to spend around 1-2 days per week on it over a period of 30weeks, which gives 720-1440 hours... so yea, I am prepared to spend a butload of time in this thing ;)

Quote: Original post by Telgin
As for specifics, well, that is hard to say. You could at least start with the obvious if you're going with a physics engine: basic movement (velocity, acceleration, momentum, elastic / inelastic collisions, rotation kinematics, stuff like that). If you get that far, then look into the cooler / harder stuff.

Another thing I just thought of, and I was thinking of starting it a while ago but never found time for it, is real world physics applies to a car simulation. At least on that subject I know I have a lot of material to read, inclusive in this website. My only fear is that is I might be doing something on a too specific subject, so that it won't be worth much in case I want to work in, say, a RPG company.. Good idea or way too overdone?