It is interesting how different things actually play out than what you put into the game as code and logic.

Here’s a list:
I found and fixed a bug – stations couldn’t be destroyed. I don’t know how long this bug was in effect, since the enemy now builds stations this bug became apparent after I ordered the Command ship to destroy a defense station of the base I was assaulting. Speaking of stations, I recently changed the health points of stations that were being built, I made them start at 0 and increase as the station was being built (prior they started with max health even 1 second into their existance). Meaning that in theory the player could destroy the station as it was being built and not yet operational. The Command ship was firing all its guns and missiles and did put a dent in the station but couldn’t destroy it (the health generation was overwhelmed the damage), so I need to change this process and stop the station’s health growth if it is under fire. Otherwise the situation is a bit silly.

Minor bug regarding ships wanting to return to a friendly base, instead they should patrol the enemy base until the player commands them otherwise.

After defeating the defenders I realized that I have to tweak a few things, I started sending in the boarding ships, and realized that it would take quite a lot of them to kill the crew of the base which at this point numbered at over 1500. I plan on putting a cap the station’s population (this is a easy way to control difficulty as well, nearer stations will have smaller cap than further away bases). To create more strategy I plan on putting in “city” stations, which will allow for a larger work force (raising the crew cap), but they are much more vulnerable to attack since they are outside the base in a station that can be destroyed. The enemy and probably players will build defense stations around these too

The base’s crew generation (it is a percentage based one) was far greater than the rate of boarding ships I was sending so I need to stop the increase of population if the station is being boarded, plus put a max on population growth instead of making it a straight percentage only deal.

I also plan on putting in different ways that the player could hurt the base’s crew, for example a station that shoots torpedos (or even meteors, why not eh?) at other far away stations and bases, if shot at a base it would kill 5% of the station’s occupants (meaning that this torpedo station alone wouldn’t be enough to destroy an enemy). Plus the torpedos will have low health so that defensive stations will have a chance to destroy them. And with the city stations I mentioned, base assaults will become much more interesting. I think this will more strategy and fun to the game! Very excited about the future.

Well I don’t feel as happy as I thought I would finally adding in the command ship. The game has changed a bit, the difficulty decreased significantly since the command ship can easily be upgraded at this current state. I’ll have to work on toning this down and slowind down this process, giving it a real cost instead of just 30 seconds per upgrade.

The truth is that the project must become a better game as a result of adding such a complex thing as I imagine the command ship becoming. It will be your base, your command center, your seat of power, your warship, and essencially the player’s physical representation in the game. Something that has to be invested in, protected and risked. So the game will need real risk and real reasons for putting the command ship into a dangerous situations. I’ll have to give the game objectives for the command ship to conquer. That means more dynamic world, better AI, and stronger enemies. At the same time I plan on implementing command ship abilities, like shields, heavy weapons, and quick get aways.

Right now the game isn’t as good as I think it was in iteration 4, the mindless fun of kill until you are killed is a bit altered with the addition of the command ship. I’ll need to preserve this, and add more strategy, more goals, more game. Oh well, I hope you come along for the ride

So I started work on converting the AI to use the new physics system. Before the AI would know exactly how to turn, if it needed to turn right 90 degrees and the ship had a capacity to turn 80 degrees a second, it would then turn at 80 degrees per second, then the final 10. Now that turning is velocity based this is a bit more complex. The AI now has to decelerate the turn or it will over shoot the target.

At first I had a bug where the AI would constantly overshoot so it looked like the AI was drunk, weaving left and right eventually slowly getting to the correct point. I’ve fixed that bug, but the AI has a bit of a problem rotating correctly to hit fast moving/turning targets.

This may not be totally bad since this solves a problem that I had for a long time, the AI was far too accurate and the player didn’t stand a chance on a one on one fight. Now that the AI can’t turn perfectly anymore and has to slow down the turn the player has an advantage. The problem is that the AI is now too disadvantaged, especially when it comes to beam weapons. Before beam weapons were clearly superior to projectile weapons, now they are inferior. So what’s the solution?

One thing that I think I should invest much time into is making the ships a bit easier to maneuver, not as easy as before, but the deceleration of a turn should be faster to give both the player and the AI more control. Its pretty neat and feels fairly “realistic” the way it is now, but needs tweaking to be more fun and arcade like, a good balance between both worlds to reduce unwanted frustration of the player.

The other thing that I can do is give the AI the same “cheat” ability that the player has had. When the player ship’s weapons were near the target, they would turn within 30 degrees or so and help the player aim. The AI might need this ability as well.

———-

The other neat thing about the better physics integration is that now there exists the “correct” momentum of a ship. Before I had a value that was most of the time correct, except when a ship would be colliding against another object, then the value was horribly mistaken, because the ship’s momentum wouldn’t be affected by the impact (under the old system, now this is under the control of the physics engine). This might not sound like a huge issue, however when it comes to aiming and AI it was big. Weapons have a built in “aimNode’ (as I call it) that calculates the correct position for the weapon to be aimed to have a great chance to hit their target. Part of the formula was the target object’s momentum and the firing object’s momentum. If this value was off then the weapons would miss. This off value would crop up a lot when objects were colliding, the weapon AI would still think it was moving and thus shots would be hopelessly off target.

So not only was the old physics bad in the sense that collisions were very sticky and ships sometimes would get stuck on each other, but in many cases weapons wouldn’t be able to destroy them, aggravating the stuck issue because the ships were basically invulnerable due to weapons constantly overshooting them.

As I was play testing Iteration 4, I watched as a large carrier turned maneuvered. Problem was that the large carrier was able to stop turning instantly. This is not the type of space sim I want to make. Don’t get me wrong I don’t want the ships to be sliding left and right either, hard to maneuver, but I want there to be some weight to them. I’ve been planning on fully integrating the Bullet physics engine for a while and started down this path again (I gave up many times).

The last attempt of me doing this resulted in a hybrid system, I would position and orient my ships manually, but when a collision happened I would give the physics engine control which would then show a dynamic collision. Problem was that doing things this way severely limited the physics engine’s understanding of the situation, because it didn’t know about the linear and angular velocity of the objects, the collision wasn’t dynamic. I would limit the physics engine control over the object for a split second because if the physics engine had control then the player wouldn’t, so I limited this to the point of impact and a few moments after, meaning that the collision would start abruptly, then stop abruptly, no fluidity. Plus the ship’s momentum wasn’t impacted by the collision.

Anyways the above system gave me more then I had before so I was fairly happy with it. But now I have a much more integrated system, now I don’t rotate an object, I apply a angular force to it, I don’t move a ship, I apply a linear force to it.

Here’s a before and after video showing the same collision happen in the two modes:

The change is subtle, but the after is more fluid and dynamic, and best of all the player has some degree of control at all times. What is left is for me to teach the AI how to maneuver with this new system in place as well. Hopefuly the benefits here will also be neat.