Engine design, components, and Papervision3D

So I just got finished implementing components to replace my inheritance based entity system. Getting used to factories and pooling has been a bit of a battle, but I feel as though I’m ready to start actual game production. My engine has fake lighting, which adds another draw call, and it slows things down a bit. Typically you can get away with 1000-1500 triangles in a Papervision3D scene without worrying about the CPU melting, but I’ve limited my game scenes to around 500 triangles, which is 250 quads or 40-ish cubes.

ShowAll, 100%, draw, compositing, and a quadcore computer.

After a little debate in FGL chat about framerate (30 looking choppy and 60 being smooth), I went and looked back on an earlier version of my engine. At that point I had lighting implemented, and was able to have around 500 triangles with a rock solid 60fps. The old engine wasn’t too optimized, and wasn’t built for garbage collection, so I went in a new direction.

Quick and dirty film grain effect

First of all, I think I’m done with daily blogging.  It’s too stressful, and I still post pretty often, so don’t worry. It becomes a problem when I have to post every day though and all I can come up with is lolcats. I’ll try posting some development related things soon. I’ll also be cleaning up the blog again at TVBs suggestion.

Now onto some game stuff.

It’s a zombie shooting game! Of course I’m going to add film grain! It makes things more gritty!

Film Grain Effect

Indie game development – What an awesome world!

OK, maybe it’s from lack of sleep… Maybe it’s because I’ve stayed up all night and morning reading from various indie development blogs. Whatever the reason, I gotta say that the indie game world is one amazing place.

Molehill Racer

I’ve recently come across some developers who’ve tried to document the Molehill source code visible in the Adobe videos just so we can get an early start. Such dedication! So awesome! From Flash to Unity, from browser, phone and even consoles… Everywhere  I look I can find indie devs working hard on their game projects, and making money to boot! I never thought that anyone could be making money with these small 1-2 man game teams until I jumped into it myself.

Going back to Papervision3D.

…Away3DLite was a fun detour, but I’m gonna go ahead and drop it and go back to my old system. I’ve tracked down my memory leak when using Papervision, so I’m getting back on that boat (I’ll probably switch again when molehill shows up).

Doh!

Papervision is faster. That’s all there is to it. It can display more objects with steady performance. Sorry Away3DLite!

Away3DLite and new recording software

Engine was stripped down, rebuilt using Away3DLite. Now I have to turn it back into a game.

Pretty huh? Too bad it's not a game yet.

Well, what can I say? Switching to Away3DLite was a larger task than I thought it would be, and now that things seem to be working, I’ll have to copy / paste a lot of my old work back into the project, and adapt it to work with the rebuilt engine. Isn’t making games fun?

Done with Papervision3D. Hate Away3D… Love Away3DLite!

Ok, so the last 2-3 posts were pretty hate filled because a papervision bug almost threw my game in the toilet. People say how much better and similar Away3D is to Papervision, because it’s a ‘branch’ or derivative. But the only similarity you’re likely to see these days is the painter’s algorithm. That’s where the similarities stop. There are several reasons I don’t like Away3D. First of all, if feels bloated. Secondly, you have to create an object to pass parameters in the constructor of every object. Come on. Strongly typed variables or GTFO.

I’m not happy about it. Away3D performance is terrible compared to Papervision3D, and the only way it can possibly be useful is if I use Away3DLite, a minimal ‘branch’ of Away3D. And oh lets see… It’s been optimized and stripped of all the flashy features. Well that’s nice.

FINE! I’LL SWITCH TO AWAY3D.

So I started to notice a little bit of a performance drop, and I could swear I smelled memory leak. I’ve been using an older version of FlashDevelop, and noticed that the newer versions came with a memory profiler! Awesome!

Long story short, Papervision3d has some problems with ViewportLayers. I submitted some issues to the googlecode and how to fix some of them, but now I have to trace down what’s doing this…. Or… I could switch to Away3D.

Shit.

Is Papervision3D really dead?

So I’ve noticed there hasn’t been too many updates on Papervision recently, and it’s understandable because there hasn’t been much out there to increase performance, even though I see away3d constantly getting updates. I don’t really care for Away3D. Away3D performance is poor, and I got started with Papervision, so I tend to stick with it. Not to mention the documentation that’s already out there. And now I see on the forums that the project has been pretty much abandoned, even with Molehill around the corner! What the hell guys? Do I really have to move to Away3d?

EDIT: Now that I’ve played with the Away3D Molehill demos, it looks like I’ll be switching to Away3D once Flash Player 11 is released. Until then, I’ll stick with Papervision because it can push more triangles in more DisplayObject3Ds. Papervision still draws faster than Away3DLite, so until Molehill is widely adopted, there’s no point in changing renderers. They’re very similar anyway, so the switch will be easy.

Overloading my game with effects

So now that I’m pushing hard to get my game out, I’ve run into some issues. I’ve had to cut a lot of features to get the game out ASAP, and this has helped for the most part. But there are some features which are pretty much finished, but not truly optimized. Here’s an image of my game with bloom, lighting, and other effects applied.