On The Wing v0.06: Castles

After a long break I’ve returned to working on the game, and here is a new update!

New features:

• Mana dropped by monsters can be “possessed” with a special spell, it then increases player’s maximum health and mana.
• Player can build a castle which serves as a home base where the player is much better protected than in the wild. (The castle model shows off my incredible 3D modelling skills.)
• Castle is also a home for mana collectors, that search for mana possessed by the player and bring it to the castle, this increases player’s health and mana even further. (Another demonstration of my great 3D modelling skills. I probably need an artist or something.)
• Sprint ability (by double-tapping forward key): player’s speed is significantly increased, but player’s mana is consumed while sprinting.
• New terrain generation code based on Voronoi cells, somewhat similar to this algorithm, but on a sphere. It, in theory, gives me a lot of control over terrain generation (for example, shapes of islands and continents), as opposed to completely random Perlin noise based terrain.

On The Wing v0.05

A new build v0.05 is available right there.

Not much changed since previous progress update: added Heal spell; all spells now consume player’s mana; many small fixes and tweaks.

Changelog since v0.04:

• Added GUI: menus, options screen, HUD displaying player’s health and mana, etc.
• Experimental support for split-screen multiplayer
• Monsters drop mana (can’t be used by the player yet)
• A lot of other changes

On The Wing Progress Report: Split-Screen Multiplayer

Last week I (somewhat unexpectedly even for myself) implemented a feature I hadn’t even considered before: split-screen multiplayer. Up to 4 players are supported, just make sure you have a big display and enough gamepads.

It turned out much easier to make than I’d thought (took me a weekend though and is still work in progress), and also very useful because it forced me to clean up a lot of stuff, such as input handling and GUI, and introduce the concept of multiple players which is the first step towards networked multiplayer.

Other significant changes:

• A lot of changes and tweaks to the input system, now it’s possible to switch control schemes between Magic-Carpet-like and traditional FPS-like controls (there is no GUI for that yet). The difference is that in Magic Carpet, horizontal mouse movement controlled rotation speed rather than rotation angle (like in first-person shooters). I prefer the former since it makes a better feeling of flight, but the latter might be easier for many other players because of its familiarity.
• Added basic HUD for players, displaying health and mana amount.
• Improved menus, now it’s possible to create a new game, then destroy it by quitting to the main menu, then create again — something that wasn’t possible before.
• Started working on the Castle spell, still a lot to do here.

A Game With A Name

My unnamed planet game is unnamed no more! It’s now called On The Wing, which, I believe, perfectly fits the nature of the game.

Planet Game Progress Report: GUI and Menus

A lot of work done this week and not much to show.

The only new gameplay feature I’ve implemented: monsters now drop mana when killed. You can’t do anything with it yet, but at least it looks nice, especially during the night. :)

The biggest thing I’ve been working on is adding GUI to the game: main and in-game menu, settings screen, progress bar when generating a new world, pause functionality and all that stuff, which is required for a proper game (it all is still work in progress). At first glance, adding menus doesn’t look hard, but in order to do it I had to refactor a lot of existing code, which is a good thing because now it’s much better organized.

For the GUI, I decided not to use Unity’s own GUI functionality (which, according to popular opinion, is too far from perfect) and after some investigation chose NGUI instead. NGUI looks like one of the best options for building user interfaces in Unity and so far I like it, though it took some time to learn how to use it.

The next thing I want to implement for the game is save/load functionality. This will delay adding new gameplay features once more, and is also quite boring to work on, but I have to do it now to save time in the long run. Save/load system is that kind of functionality which, I think, benefits from being implemented as early as possible.

Planet Game v0.04

The biggest change in this build: I’ve implemented an aggressive AI for monsters. They’re now switching between two behaviors: Wander (flying around randomly) and Attack Player, when the player gets too close to them. While attacking, monsters shoot fireballs at the target and “dance” around it in an unpredictable way, which makes them harder to hit. Now these monkeys are truly dangerous!

While this behavior doesn’t look like something big and complicated, it required writing quite a lot of code to implement two systems: steering behaviors (which are particularly difficult on a sphere) and a goal system, which allows characters to have different goals and goal evaluators that are used to find the most important goal at this moment. Now it should be possible to implement more complex behaviors when I need them.

Other, smaller changes:

• Added shortcuts to change mouse sensitivity (“<” and “>”). Some people complained that control scheme is difficult, and one of the reasons might be that I tweaked them for my sensitivity settings, which are higher than normal. However, it also possible that copying flight mechanics from Magic Carpet wasn’t that a good idea.
• Added player’s health status on the screen. Nothing happens, though, when the health reaches zero.
• Added help screen (toggled by “F1″) with the list of controls.

Planet Game v0.03

Last week’s progress on my new unnamed planet game:

• Added Fireball spell; rewritten spell system, it is much easier now to implement new spells
• Added basic code for 3D steering behaviors that will later be used for implementing NPCs’ AI
• Added flying monsters that can be killed by the player using Fireball. Monster models are obviously placeholders, I used the standard monkey head from Blender. They can’t fight back yet and just wander randomly, so killing them gets boring after a while.
• Tweaked player controls
• A lot of other small changes

And here is a screenshot of the new Evil (but harmless at the moment) Monkeys (as I said, the model is just a placeholder)!

A New New Game Idea

Yes, it happened again. I’ve postponed that voxel engine game I was working on because I want to try something else. I always liked an idea of making a game where the world is a real spherical planet, a little one, 1 km radius or so. Last two weeks I’ve spent working on a small prototype (I used some code from the past, but rewrote most of it anyway), inspired by Magic Carpet, one of my favorite games. The prototype now has a real spherical heightmap-based terrain, which can be modified in real time (there are three different “spells”). I’ve even recorded a video to show it off. :)

Voxel Engine, First Results

It’s been a month since I started implementing a basic voxel engine in Unity 3D for my game(s). What I have now:

• Chunked terrain, editable on-the-fly (adding/removing blocks in the game)
• Mesh generation
• Optimized collision meshes which are generated separately from the main meshes and have much smaller number of polygons
• Colored lighting backed into voxel data
• Simple shader which supports texturing (I’m using only one texture at the moment) and both backed and normal lights
• Shooting (also destroys blocks)
• Objects that represent light sources (those light panels in the screenshot)
• Modified chunks are rebuilt fast enough so that there isn’t any noticeable lag
• One screenshot :)

Isn’t that much, but at least something. Now, after I have a basic engine, I can focus more on other stuff: generation of more interesting levels and adding some interactivity, such as doors, light switches and so on.

Why Voxels?

So, this is my second serious attempt to create my own game. The first one seemed promising at the beginning, but eventually I became somewhat disappointed with it, for different reasons, mainly because of too big scope for a first game. (Some day I might revisit it though.)

Now, I have a high-level understanding about what this new game is going to be, but I haven’t yet decided what subset of ideas that I have I want to include in that game. Will write about it later.

What I definitely know, is that I want to use a voxel engine for this game; the simplest, blocky variety of it, not Marching Cubes or something. (Don’t even mention Minecraft, you! I’m not going to implement neither mining nor crafting!)

Why voxels? Well, for a couple of reasons.

First, I like them as a programmer: it’s a good choice for randomly generated areas, which I need for my game. Heightmaps are dull and their use is very limited when creating indoor areas, caves etc. Marching Cubes are nice, I played with them a bit and managed to get a working implementation (see this great paper if you’re interested in such kind of stuff), but it’s way too much work compared with cubes, I’m not ready for that. Cubic voxels, however, are relatively easy to program, and there is a lot of useful information out there about various implementation details. For example, this thread at Unity forums has been very inspirational.

Second, I just like them. They can even be beautiful when done properly: take a look at Cube World for example, It’s amazing!

Another reason is art assets. I’m going to design areas and create all models myself, but I’m not an artist, and while I can learn how to use traditional 3D modelling tools such as Blender, it will take too much time with questionable results. On the other hand, I spent some time creating models in voxel editors (I especially liked Qubicle Constructor, I think I’m going to buy its Unity Edition) and I liked both the process of “drawing” in 3D and the results I’ve got. I could use voxel models with any kind of engine, of course, but voxel terrain is the most natural environment for them, obviously.