Stage 3 was the first area planned. The same day I had the thoughts of making a Hunters game, I began writing down ideas for a city stage that would have you traveling through sewers and then into a burning building. The detailed changed a bit, but the overall idea stayed the same.
The scrolling area was something added in a bit later to help work the stage into the story, as well as provide a slight gameplay break. Even though minor, it was also the next step to creating a fully scrolling area, even though the enemies stay away when the background is moving, it was still nice to get a feel for Carina moving through an area, and platforms moving opposite her.
MS Paint version of the sewer area. The flow of the sewage is all noted on it. Being one first major cross-room elements, it helped to know where it would be in advance. In challenge mode, walking in the sewage slows you down. This is something I wasn't able to code easily when I first made the stage. Later on though, when I added sound effects for landing on a platform, a system was added let the game know what each platform was "made" out of. With this in place, adding a slowdown effect was a cinch, but I kept it to challenge mode only, not wanting to risk making the regular stage too difficult.
The sewers were originally a bit longer. On the map above, three rooms were added where the yellow arrow is, to allow you to reach the lower section faster, cutting about 6 rooms on the way to the boss.
On the trip up when you return to this room, you can go left or right, but I suspect most people will shoot the console, and then towards the path it opens, and the rooms that they haven't seen yet, reducing repetition.
An unused idea for this area was that after killing the mini-boss, either the water or Carina herself would become electrified, so you would have take a slightly different way back up to the top, with the added challenge of not touching the floor.
This area was a lot of fun to design. It's more puzzley than the rest of the game, and was tricky working in the extra elements of obstacle and console into the single screen layouts.
The healing consoles were added later on as a way to lead players the correct way in confusing rooms. Since there is no way to gain health in parts with no enemies, the familiar yellow glow becomes extra inviting.
Stage 3 initially saw a huge inflation of enemy HP values, that I later toned down significantly. In stage 2 the rocket could kill almost anything in one hit, so to make the charged up handgun seem strong, I set the HP values of a few enemies way too high. The bouncy blobs and large rats had 10 HP, and the cockroaches had 12 HP. For reference, a rocket only does 6 damage, and a fully charged handgun shot does 7.5. I thought to take down the enemies you could do a fully charged shot quickly followed by a regular shot, but that didn't feel as satisfying as just a single charged shot.
The blobs and large rats were lowered to 7.5 HP, keeping the feeling of the rocket not being very effective on them, while letting the charged handgun kill them in 1 hit. The cockroaches were lowered to 6 HP, letting the rocket not totally suck in the first half of the stage. Since you never get any stronger weapons after this, enemy HP values stay pretty consistent for the rest of the game, except for very large enemies which shoot up to 18-24 HP, at which point the rocket and handgun are both equal in kill speed.
Having used pretty much every simple enemy design in stages 1 and 2, I had to start thinking outside the box for new enemies. While the blob is a fairly simple enemy, the big innovation here was adding in code to allow them to reverse their personal gravity. Even though they're one of the only enemies to use that trick to get the drop on a player, having the code around simplified placing enemies on ceilings in later stages.
One persistant idea I wanted to work into the game, but had little success with, was having obstacles that couldn't be destroyed. I felt in a lot of places, allowing the player to kill everything and then walk to the next room was too easy, and the sewer area was the worst for promoting this style of play. The ceiling slimes were originally killable, but after watching a friend play through the game, they ignored them, not knowing you could shoot them. Realizing that you never needed to kill them, I made it so you can't kill them on normal and hard, allowing for some rooms to always have some minor danger to them.
The slime monsters were another enemy specifically designed so that the stage would still have some projectile shooting enemies. At first they would only shoot straight left and right, but since they were only placed on floors with sewage, they weren't very effective at hitting the player. Eventually I decided to upgrade them to a 3 way shot, which made them much more effective.
The sewers in general was tricky with enemy placement, as in addition to keeping the fish and slimes in the water, I wanted to keep the rats and cockroachs out of the water when possible, as it looks weird with them. I wanted to have an on the walls enemy for stage 3, but could never ended up thinking of anything that would work.
An unused boss sketch. I'm pretty sure it was intended to be the Stage 3 spaceship boss, before I decided to just use a smaller ship, and have it match the ones you'd fight later on in the shmup area.
All of of the rough boss sketches were done in December 2008 when the power went out for almost a week and I had nothing better to do. Even though I was only up to stage 2, I knew what I wanted for a good number of upcoming bosses. This sketch of the slime boss is very close to the final version. And yes, my handwriting really is that bad.
More detailed version of the slime boss. A few of his attacks are based off the old idea of having the water become electrified and dangerous.
Rough sketch of Draco. He didn't have enough attacks yet.
Final version of Draco. He's definitely one of my favorite bosses in the whole game. The randomized attacks make it a much different encounter than anything else before, and I was able to really go all out with the attack animations, something I wouldn't have been able to do at all a few years earlier.
One thing I noticed myself doing in the Project Inthri games is having a stage 1 boss with multiple hitboxes, then getting lazy and not doing that again afterwards. Since I like multiple hitbox bosses, it was a very conscious decision to make both big shooter bosses in that style. Draco himself was sort of envisioned as one of those big battleship type bosses, where every tiny part has its own turret. The fight starts off very difficult, but as you begin wearing it down and breaking parts off, it becomes much easier.
There is a minor mistake with this scene here. Originally Carina was going to look up and see Avior crash into the building, but I decided to change it and just have the building already be on fire. The problem is that Carina didn't come here chasing after him, so should have no idea that he's nearby. Maybe her magic watch told her instead?
The game itself was made well before the cutscenes and script were done, so during the Draco fight I thought Avior would still be passed out in his ship. In the final version though, he's awake and just sitting there motionless, which looks kind of weird.
This is the first of a few videos I made to show off different bugs that were found in the game. In this one, a single platform being too low allowed Carina to jump back up to an area that's supposed to be inaccessible.
This was one of the most frustrating bugs in the whole game. Enemies that turn on the edges of platforms have a lot of special properties, their hitbox doesn't work quite the same. If you can knock them out into the air in a way the game doesn't expect, weird things can happen.
Another bug involving enemies that turn on edges, if an enemy died, then hit the edge of a platform on the exact frame his momentum ran out, the turn animation would take priority even though the enemy was dead, leaving a ghost behind.