What kind of game I want to make...and what kind of game I can make?
Take my case, for instance, and my latest effort. Sure, I would love to be able to make a modern FPS, with all the bells and whistles, employing next-gen graphics, vast content, etc etc. Sadly, of course, this is close to impossible for a one-man-team, that just has hopes of expanding when the time is right. However, I believe I still can make a fun and satisfying FPS, with all the limitations that my situation imposes. In fact, instead of fighting against the limitations, my plan is to accept, embrace them and make them an integral part of the game.
Graphics, for starters, as we know, is for good or for bad a big factor in this genre. Now, I'm not deluded enough to think I can even compete with the monsters out there(Unreal3,COD,BF3,Crysis,Rage). Heck, I don't even think I can compare to,let's say, previous-gen engines(Doom3,HL2). The problem lies not so much in the technology available(after all, I could ditch my little pet engine and use UDK or the newly-released CryEngine3 Sandbox if I want), but in the content. Those games have terrific engines, sure(from a visual standpoint, at least), but they look this good mostly because of terrific art. I have invested some of nice Euros(thanks to a new job, I can now do that) to buying some art, but let's face it, it's not of AAA quality.
So, if the graphics are surely not going to be next-gen, non current-gen, and probably can only begin to compare to previous-gen, what are am I even talking about? Well, I decided that all the comparing is not worth the effort. The game will be what the game will be, and that is the best that *I* can make. And here come the limitations I was talking about. Based on those, I decided for some general guidelines, for graphics,gameplay, but also about the 'marketing' itself. Those are, in no particular order:
1) The game will be freeware. I used to think I could maybe pitch a game of mine, if I ever finished it, to Steam for a low price, say $9.99. But now that I think of it, even that is maybe too much to ask for a game competing in an overpopulated genre as first person shooters. My purpose, after all, is to finish a game and have as many people playing it as possible. Build a fanbase, so to speak. If I can achieve that, then we can maybe begin about commercially pitching any sequels or subsequent games. What I'm a bit worrying about is whether that will enable the 'if it's freeware it's crap' mentality that many have. Comments on this?
2) It will run on older hardware. If it's not graphically that impressive, it should at least compensate in that department. Targetting an audience that can't afford(or want to afford) the next Crysis may prove to have some pleasent surprises.
3)It will mostly feature outdoor, natural environments. Now, this is due to the limitations of the project. Comparatively, nature scenes a-la Far Cry as somewhat easier and cheaper to make. An urban environment, for instance, needs lots of content. Not that nature scenes are light on content, no, but it is my impression that you can achieve more varied results with less effort.
4) The gameplay will be mostly old-school. This is mostly due to my personal preferences. I'm starting to get tired of the same pattern on nearly all shooters- cover, regenerated health, limited weapons at a time, checkpoints, etc etc. So the game will be a return to the roots, and will speak to those that seek a new fresh shooter that plays by the old rules. So it will have a health counter and not regenerated health, powerups and healthpacks, and probably quicksave/quickload capability. At the same time, I will throw in the mix some slight RPG elements. By taking down enemies and completing objectives, you will gain points which you can use to 'buy' some upgrades, like bigger health bar, more stamina, speed, agility, aiming.
All that said, the game is still in its infancy. I worked this week and finally got shaders up and running. I implemented basic terrain texture splatting using a simple shader, and the immediate plans for next weeks are:
1)Normal mapping, parallax mapping.
6)optimize the particle system
There isn't anything on that list that I haven't implemented, on way or another, on past projects. I especially expect that vegetation and shadows will really live up the scenes. For vegetation, I will probably use a technique I had used in an older project of mine: Basically, you create a patch of grass, and you repeatedly render it around the camera, ofsetting the height of each billboard by sampling the heightmap in the vertex shader. Using also a vegetation map, you can define what areas have vegetation and at what density. Here it is, as it was implemented on one of my previous projects:
I got pretty good performance with that on my old NV6800, so I expect that things will be even better in my 9800(which, by todays standards, is low-end anyway).
Ok, that's it...I close with 3 screenshots of the game and one of the editor(which is not WYSIWYG yet). Any comments are appreciated!