I thought about the extension and I will do something, get a bit loser to working. But I wasted most of today, working on a way to run the graphics through OpenGL instead of Core Graphics. That could give both portability and much better performance.
I probably won't finish the OpenGL based version, but I have also implemented a queuing system for game objects that reduces the load from objects far away. That allows me to enter a huge number of objects to populate the world. Now I just have to make a few more kinds of opponents and that kind of things.
The input queue works just great, but I have to make two more, one stack at least side where I store all objects that are too far away from the player to be interesting. Then you can go back and forth, everything will remain, and I can have thousands of objects costing only memory.
In the past, I would have used a linked list for this, but now I am using dynamic arrays instead. That allows me to do it with fewer memory operations (granted that I expand the list with many items at a time, which I do).
My "large world stacks" were really easy, worked almost at the first try. So now I can happily add a ton of game entities. The only risk is that it slows down the game start, but only that.
Now, what more enemies should I add...?
Well, it is still just a first draft, but 0.0.3 at least has a some more of those features the full game should have.
- Many upgrades work.
- Aeroplane not totally impossible. But not really good.
- A bit more flexible on coins and cannons.
- Jumping works.
- Vehicle hit points.
and a whole bunch of other changes... but the sound of the "spawners" is still annoying, no engine sounds, and no winning (last sections are unpopulated). I know what I wanted to do to make a neat ending, but I don't have the time. Heard that before? Same things with my OpenGL acceleration/porting interface, it would be great, but so much left.
A few last fixes, populated more of the world, planes can shoot...
Bookmarks