This blog post is my journey into Game Off 2023 by Github and Lee Reilly, started on November 01 and ended December 01. It's my very first big game project, I am so exited! I will update this page every day, telling what I did and the expectations to the next day.

#Day 01

I was so excited to participate. I remember only making one game in Javascript very simple, three or four years ago, I think using P5, It was a runner game, like the Google's Chrome Dinosaur Game, very nice, but I think something more complex.

This day was the great theme reveal, which was Scale. It was very nice, very embracing. The first moment I thought in making something like Cuphead, a boss battle rush themed as pirate or other theme related to sea that I can fit scales.

Thinking more about, a boss battle rush would be very difficult for my first game because I need to create a lot of mechanics, and in one month maybe it's not possible. So I decided to go to scale, an increasing stuff. I thought in games like Vampire Survivors, a bullet hell game. Maybe I cannot do a lot of enemies variations, but I think I will go if this.

#Day 02

This day I was searching for a game engine. As I only code one game, I do not have an expertise of how game engine is better for my choices, so I take a time to study some of the engines recommended in the Game Off 2023 page. I started in Unity, It's a very good engine to design first games because it's simple and intuitive.

I also look at Unreal, with a very good lighting system and incredible stuff to handle 3D games it's probably the best engine out there for someone trying a next gen game, but for me that is my first game, and will probably be a 2D game, Unreal would be kill a mosquito with a bazooka, and for this, Unreal will not be my choice.

The last engine I tried was Godot, a very good engine open source that you can create 2D and 3D games, but I think the greater part is for 2D games. Godot has a scripting language called GDScript with a very simple syntax inspired mostly by Python, a programming language that I used a lot on my side projects.

In the end I chose Godot, mostly because GDScript is so similar to Python that I had the better experience of the three ones. Godot is also intuitive, and I think my challenge in the next days will learn everything about Godot and create the game.

I had a little extra time in this day, and started to learn about Godot using the Godot's fundamental's playlist in the Game Dev Artisan channel.

#Day 03

This was a not good day. I mostly study how to use Godot, I learned to create collision and pickup items, but I tried to refactor my files and I delete my Player with all the animations and logic already finished 🥲🥲🥲. I stopped and I hope tomorrow will be a better day.

Well at late night I thought better and decided to start a new project, this time, with a better preparation, and clarification of the ideas. I started searching in the assets I downloaded at Ichio.io. I found a very good collection with free license Ninja Adventure, it has a full set of incredible pixel art with event dialog faces that I can use in the buy menu. The next step was think of what type of game I wanted and maybe, a tower defense it's really nice and I can reuse some code to other enemies. Other point is the map, I just need to create one for this first time.

To get started I search in the assets pack and decided the towers. Of course I can have a lot of towers, but the game is scope is to be small and fishable until December, so I cut and decided to create only 6 towers, each one being unique and having a upgraded version.

Name	Damage	Sprite normal	Sprite special	Weapon
Knight	area, close
Mage	single, far
Support	None (gain coins and life)
Ice man	area, far (Slow the enemies)			Ice ball
Defense	None (Stops the enemies with barrier)			Stone Rock barrier
Summon	single, close

And when thinking of enemies, I would like to show the first boss early because, the objective is to create a simple game, so I decided to separate into 20 levels and each level will have a new enemy. Maybe I will change this in the future and just have 4 levels with a longer duration each one with a boss in the end. I don't know, and we will see what I end doing. For now, the level are these:

Level	Slot	Slot	Slot	Slot	Slot	Boss
1			x	x	x	x
2				x	x	x
3					x	x
4						x
5
6			x	x	x	x
7				x	x	x
8					x	x
9						x
10
11			x	x	x	x
12				x	x	x
13					x	x
14						x
15
16			x	x	x	x
17				x	x	x
18					x	x
19						x
20

The main idea is that each tower can be good/bad to a enemy type. For example the Knight tower is only able to damage near enemies but can deal area damage, so It's good to enemies that come in groups(Mostly land), on the other hand the Mage tower has single shot and far distance, which is good to fast and single enemies(Mostly air).

#Day 04

This day I was little busy playing tennis with my brother and my girlfriend. Was the first time we played, it's nice, but tiring. Anyways, in the spare time I worked at creating the project on Godot (I know it still has the songs and tiles, but this I choose later). This time I looked a lot in the Godot Documentation therefore I don't have much to say, I mostly just studied.

#Day 05

Today is Sunday, so I have all the time to play around with the game building. I started to follow a Youtube series from Game Development Center, it's a bit out of date, but we can get a very good idea to how tower defense games are.

The first classes is to structure the project and make the first configurations on Godot. He said that we can reuse this to future projects.

First, the folder structure. It is separated in 6 folders, the first one being the Assets, that is branched into Environment, Icons and UI, everything auto explaining. Inside the Environment we have Props(Files with one asset per file) and Tilesets(Files with two or more assets in). The Licenses is where you store license of fonts, assets and everything else. The Resources is objects that will be reused on project. The Scenes is where will be store all the objects(Classes) created. Lastly we have Singletons, objects that are independent from the rest of project.

Other thing I did is the first map, I finally learned about tilesets. Following the documentation and the video of Game Development Center I finally make the first map using some scalable techniques like make terrains.

Terrains for now is just used to create roads, that using assets Godot can automatically find the right asset to that road, this way I don't need to find the right asset to that corner on the road. I followed this tutorial by Maker Tech.

The map has still really poor flora and I want to trees like terrains, to be more efficient in creating forests (If you pay attention, there are some tiles with bugged trees). In general was a very productive game. If I hope I can keep up this rhythm.

#Day 06

This day I was busy with some stuff and could not make any progress 🥲🥲🥲

#Day 07

I know I should give more attention to the map, but everything seems kind weird. The scale doesn't seems right. That's because I need to confirm that the scale is right to create the map based on. So I decided to create the first tower of the game, thinking mostly in adapt the scale (Spoiler I did not have the time to play with the scale).

Anyways, I started with node2D and then put the sprite inside, I don't know if this is the best practice, but for me, giving a wrapper of node2D makes me feel more organized. The sprite for now will be the idle:

One thing is that it has right, left, bottom and top sides, which gives me the necessity to turn the sprite as the enemy changes it's direction. I started trying the sprite to follow my mouse, after many tries, I came with a simple code, but I happy that I can understand everything from this code. 😁

pythonextends Node2D
 
var direction = Vector2(0, 0)
 
func _physics_process(delta):
  turn()
 
func turn():
  var target_position = get_global_mouse_position()
  var character_position = global_position
  direction = (target_position - character_position).normalized()
 
  if(direction.x > abs(direction.y)):
    get_node("Knight").set_frame(3)
  elif direction.x < -abs(direction.y):
    get_node("Knight").set_frame(2)
  elif direction.y > abs(direction.x):
    get_node("Knight").set_frame(0)
  else:
    get_node("Knight").set_frame(1)

#Day 08, 09, 10 and 11

I am gonna group all these days together because I really did not have a lot of time to progress in the game and the little time I had was trying to make the map looks a little bit better. It was really hard, not because the sprites, they are awesome, it's because of me, I did not have and experience grouping those tiles together in a pleasant way to the eye. I don't know, I was looking for a diverse map, having the flower, ice, rocky and forest parts, but I don't know, something looks out of place.

#Day 12

This day I created another tower and rescale the knight already created to fit better in the map. It was very nice, I learned about scaling and import pixel art in Godot 4. I also put a base in tower to better highlight on map.

The code part was very easy, following the tutorial Game Development Center I just imported the logic from a generic class.

#Day 13

It was a very short day. The only thing I could do was the main menu. I was following the video of Game Development Center, but in Godot 4, some methods have changed, main thing is to connect now is deprecated and we need to use something like [Node].[Signal].connect([method]) instead of .connect([signal], [object], [method]

#Day 14

This is one more busy day, I just make some initial HUD, following the same tutorial of Game Development Center. It was kinda simple because this hud doesn't do anything else, it's just two images.

I used the assets of Ninja's Adventure again, they are just wonderful, I just need to make them smaller, but this is problem for other day.

#Day 15, 16 and 17

These days I found a very big problem 😓. At day 15 I was intended to do the fifth part of the Game Development Center Godot's tutorial, but there were a lot of different things I did in the project that I was getting the consequences. Like the resolution of 1280x720 is too big for a 16x16 assets of Ninja's Adventure, so I tried to add scale and messed with everything else, now, 4 tiles were supposed to be just 1, and the collision system was not working and I knew I needed to make a work around that I would not like.

So after trying many strategies, I just stopped 😕.

#Day 18

I was so frustrated because of bugs that I decided not think about it today.

#Day 19, 20 and 21

Even if I was frustrated, I have not given up, and would make the game from zero if necessary. And so I did, not everything, but I can guarantee was a lot.

One of the many problems was the scale, resolution and sprite sizes. The problem I had was a pixelart game with assets of 16x16 that was running on a resolution of 1280x720 which are 1280/16 * 720/16 = 3,600 tiles to fill the entire screen. Think about it, every sprite is 16x16, even the towers which in the 3600 tiles, max 10 would be towers, of course this will cause a strange sensation, it's not in the right scale. To solve this, I scale some arts, like the towers and bases, this way, a tower had 1 tile size, but the path was 3 tiles size. You can imagine the problem when making collisions.

To solve this, I changed the Viewport width and height to something way smaller (320x192) and put the Window Width and height Override to 3 times of the Viewport size. This way, the game will have a resolution of 320x192 and will open in a window of 960x576, making all the scale for me. The game now doesn't needed to scale some arts, everything from enemies, path and towers will be 16x16 and just one tile. I did a small change on the Ninja's Adventure' assets because some tiles were way to big for me, and then I made them smaller (Logs and a type of grass).

Now using a resolution of 320x192 gives us 320/16 * 192/16 = 240 tiles only. It may look way to little, but is not. I feel I can make better maps with this size.

Of course this brought me a lot of more problems 🥲. All the assets in the scenes were with wrong sizes and so I went in every scene fixing the bugs (towers, map, ui, menu, everything 🥲. If you are starting as me, just make sure the scale is right 😅).

In this process I redid the menu, ui and the map. Here is the final result:

Note: I leaned how to animate some sprites in the map 🤩🤩🤩.

#Day 22

Now after all that remake I finally can move on to other features. In this day I finish the selection and collision/building tower system. There were some changes to make from tutorial because Godot 4 has changed some names, but essentially is the same of the Game Development Center's tutorial.

I also did a very simple thing that I was supposed to do in the first time starting this project 😅. A git repo.

#Day 23

Now I can see a game under construction. This day I make a range indicator to tower following Game Development Center's tutorial #5. This range can be configured in a separate file, this way the towers can have different sizes of ranges (Of course this file also can be used to configured damage, rate of fire, cost, etc.).

#Day 24

It has finally come, the day I built the enemies. A enemy in a tower defense game should not be a problem, because it's IA is just- "Follow the path", and so I just need to configure the path and make the enemy follow it. Can easy be done using Godot's Path2D.

My problem was not configure this path, but configure enemy's animation, well the Game Development Center's tutorial #6 uses a top down view asset, as Ninja's Adventure' assets uses side view assets, so just rotating the sprite will not work in my case.

I did by calculating how the position is different from the last frame and give the right animation. Solving this made my day 😊😊😊. So Happy 😁.

GDScriptextends PathFollow2D
 
@onready var _animated_sprite = get_node("CharacterBody2D").get_node('AnimatedSprite2D')
 
var speed = 0.5
var last_position = Vector2(0, 0)
 
func _physics_process(delta):
  move(delta)
 
func move(delta):
  set_progress(get_progress() + speed + delta)
  var current_position = position
 
  if(last_position != current_position):
    if(abs(last_position.x) < abs(current_position.x)): # Going Right
      _animated_sprite.play('east')
    if(abs(last_position.x) > abs(current_position.x)): # Going Left
      _animated_sprite.play('west')
 
    if(abs(last_position.y) < abs(current_position.y)): # Going down
      _animated_sprite.play('south')
    if(abs(last_position.y) > abs(current_position.y)): # Going up
      _animated_sprite.play('north')
 
    last_position = current_position

#Day 25

As I solve a lot of problems in the last days, I lost my track of progressing the game creation, so this day I finally had the sense of progression. I did a lot of things. First of all, I make a start/pause/fast forward system. Very simple, just followed Game Development Center's tutorial #7 and make some simple pixelart 16x16 sprites, because Ninja's Adventure didn't have sprites for arrow, double arrow and pause.

Second, the tower need to look at enemy and not my mouse 😅 I followed the most part of the Game Development Center's tutorial #8, but seeing the comments, we could improve the tower tracking.

In the video, Stefan select the enemy with most progress in the path (Which is very good, because more progress in the path, means more likely to complete the path), but this gets a problem when using more than one path. Think with me, if we have two paths and enemies in both that crossed the same amount of pixels, the tower will get confused with enemy is the right to pick. The best choice will be get the enemy with the largest progression and in the shortest path.

For example if we have two paths, one having 450px size and other having 200px size, there are enemies in both paths and the enemies have crossed the same amount of pixels, like 30. The progress of the first path will be 450 - 30 = 420, and the second will be 200 - 30 = 170, if we get the largest number, the enemy on path1 will be chosen, but can agree that is more likely to enemy on path two finish the path, because only need to cross more 170 instead of 420 in the first path. So that's why we need to get the minimum value.

After this though I made the damage system. It was very simple, the only thing I did different from the tutorial was that there, Stefan put a heath bar above the enemy. In my case, I changed the enemy color based on heath percentage like: heath > 70% = Normal Color; heath > 50% = Yellowish; heath > 20% = Orangish; else = Reddish.

The last thing I did was a shooting mechanic. In the tutorial, the towers shoot the same projectile, but in my game, it needs to be different so I abstracted the code and make a muzzle and impact animation for each tower, so thats in the future I can add a projectile to certain towers, like the Mage one, and for the Knight it does not need this. I will think more about this tomorrow.

The result:

#Day 26

This was a very simple day I just make the system of health and coin, there is no much to say. I just make some signals that are emit when enemy is killed or its complete the map's path.

#Day 27, 28, 29 and 30

As the voting day is coming, I decided to focus in finish the main mechanics instead of adding new stuff, so I decided to make the last updated in this article grouping all the stuff made in the last 4 days.😶‍🌫️

Well, first of all, as usual I went through the game and noticed 4 things to do: The game need waves counter, more enemies, about page (Where the credits of me and Ninja's Adventure will be), and for last, there are some bugs to take care.

The first thing I did was about the waves. This time I would finally make the balance between towers and enemies (I can say to that I though would be easy, but not, this is a very crucial part that can make your game look good or terrible). Was very hard to do that, but I made. I add a simple wave counter, and made the waves into an GameData dictionary so I could easily modify. I add two more screens, one for Game over and other for Game Win. And of course, more enemies. They are just an extension of the main enemy's script, with some values and sprites changed (In the future, I want to make this less manual, to add even more enemies, creating a scene in usual way is no good). In the meantime I noticed that the game was very short, so I add a Infinity mode that would spawn random enemies (In the end, this was supposed to be a very challenging mode, but proved to be not so difficult, in the future I will change that).

The about page was very easy to make. It's just a simple new scene. The more complicated part was to create a Game over and Game win because these two scenes need to be over game's scene and not in root. But it was ok, and simple to do.😊

In the bug's part, there were a lot linked to the wave system. I don't know, but making all that logic in just one script just made it very confusing. This is one more thing to change in the future, a better game's script, I don't know, maybe just separate the logic of spawning enemies, controlling the waves, building towers and enemies path in a separate script would make so much better.

It had a bug in the fast forward too, but I could not had time to fix this, so it will be a future issue 😅.

You can play the game at Itch.io. Leave a message to me. I will be very happy to talk to you 😊.

#Final thoughts

Well, what I learned with all this?

This was my first experience making a game in a professional way (In some pro engine), so because of this there was many days that I was unmotivated, even if I came from web programming, coding in game, it's harder (Maybe that's because I did not knew the language) and that made me a little frustrated, but as I was going, never giving up, I saw a light in end of the tunnel. Some stuff that was hard to understand like what is a scene, became very easy and I could make all the changes that my imagination could have. I really like the experience.

Well, there was many thing I could not had time to develop. If you see the first days of this article, you will understand that, I planned a lot of things, the scope was too big for a first project that needed to be made in just a month, but that's ok, I plan to add these things later, I will finish the game as was proposed back there.

For last, thanks everyone who read this article. It was a very difficult month, but I could say that I very happy and proud about the results. I see you in the next article. Have a nice day. 😎✨✨✨😎. dt