the sad thing is that this will probably be the most hardware intensive thing I've ever run on my computer.
also particles
Hopefully you won't have to worry. During development, I will be hugely biased... I was able to buy one of those rice-burning gaming rigs (Dell XPS 710) from work for a fraction of its original cost, so it is bad news for anyone else if my framerate is 60 fps. In the closing days of the compo I will try to keep it real and play the game on older hardware I have lying around at home, and pray that it runs well. It's a little sad that removing textures, keeping poly count low, etc., really doesn't give you a big performance boost since graphics cards are purpose-made to draw tons of textured polys to the screen. You get a bigger speedup by lowering the number of batches sent to the card and lowering the amount of CPU processing related to managing graphics (i.e. particle systems).
I wouldn't be too concerned about performance in the game. I will reserve judgment until I get the screens in the cockpit rendering real-time displays, but as for things going on outside the cockpit, you will hardly be able to see any of it, and the particle trails are really just to improve the otherwise boring visual presentation. You can expect to launch a missile or two, see it trail off, and then see a small strobe-light "poof" in the distance. Otherwise, the only other excitement is likely to be your ship's thrust contrails. I'm in the process of implementing that right now, so I will try to fraps it or post a playable demo.
I actually have a reqeust for some advice now...
I was planning to make the game all about firing and intercepting missiles in time, such that you wouldn't want to engage too many enemies or you wouldn't be able to shoot down all incoming missiles with your laser. This is still going to be true, but I think it might be sort of boring if you can easily see all the enemies in the area and they can just as easily see you. They could also shoot down your missiles, so you'd probably have to just shoot enough of them that they couldn't intercept them all. It's bad to evaluate the fun of something before you've even playtested it, but I feel this might get boring or repetitive too quickly.
I think an easy way to improve gameplay would be to complicate the radar of the game by introducing 'active' and 'passive' modes. When in Active mode, your radar range is very long, and you can see absolutely anything in that area. This also means, however, that anything else can just as easily see you. In passive mode, your range is the same, but over a certain threshhold (a fraction of the total range), you will only see targets that have active mode on. As you'd expect, on enemy radar you no longer show up in the "active" range, but instead only in the "passive" range, where you can see them as well. Missile guidance systems, however, are always in active mode, so you would always be able to see missiles, but if you aren't careful they will be launched too close to you for you to be able to react.
I could simplify this even further, and give the game a vaguely "sub hunt" feel by keeping it in passive mode all the time, but letting you send out active "pings", since in practice you will probably want to be doing this anyways. The "Earthsiege" series of mech combat games have this active/passive difference and I felt it was a fun element in those games.
From a sci-fi perspective, it is actually probably harder to spot ships in space than you'd think. When targets are thousands of kilometers away, only about 300m long, and could be in any direction you could plot on a sphere, the area of that sphere's surface you have to scan is enormous. However, ships are most likely going to stick out as hotspots of radiation and reflected light on a background of cold, dark void. I could balance out both of these theories by assuming that at a certain distance, ships are pretty easy to spot from the various waves they emit/reflect, but farther than that it becomes much harder and you'd need to start shooting out waves to see what reflects back.
Anyways, I will get back to the more pressing matters of actually getting stuff to explode and the screens to display stuff. I think I've actually got a method that will work, but I've got to wrestle more with render-to-texture since my first attempts have failed miserably.
Community
