Un-Glitch Launch

Hi all!

I'm in the process of finalizing Un-Glitch; a 2D action game developed in the course of the lowrez jam 2023. Since I currently have a near-finished game, I thought it would be a good idea to release it early and gather feedback before the jam ends. In this devlog, I will give a brief introduction, as well as some insights into design limitations and decisions.

WHERE

Currently, I am making the game available as a web application that you can play within itch.io. Executables with smoother framerates will be available for download once the jam ends and I have a definitive version.

DESCRIPTION

You enter a simulation and need to fix randomly appearing glitches by touching them. Upon fixing 10, you succeed in effectively a "debugging session" and the stage ends. There are also glitches that create dangerous electicity projectiles. Getting hit by these redacts from the number of fixes, until you also glitch and forced out of the simulation.

Along the way, you can run into buildings that provide various useful effects.

The game keeps track of a local highscore for each simulation, where the score is the number of seconds you took to reach 10 fixes. You can thus challenge your previous best score. The simulation selection and its contents look like the screenshots below. When inside, you can see the elapsed time on the top-left and the current number of fixed glitches in the bottom right. The "1+" is an effect that will make the next glitch fix count for one more.

BUILDINGS

Throughout the game, you will find various buildings, all of which can turn to glitches.  I am showing here an overview, but the game lets you gradually become familiar with them in a more reasonable pace. Pixel art is CC0 and made by me, so you can reuse it in your own projects too. I am not a good artist, but it was fun trying to make simple designs anyway. I chose a perspective that has minimal perpendicular information to keep my design basically topdown but have enough information at this resolution to understand what's there.

	First are simple houses in the simulation. They are everywhere and have nothing special going for them.
	Granaries grant double the next acquisition, be it score points or other effects. They can turn only to score-granting glitches, which lets me create small pockets of expected upside during level design.
	A simple barrier that opens and closes. It will hurt you if you are standing on it when it closes.
	Get swole real quick! Gyms grant you fast movement for a short while, during which you get to ignore projectiles.
	Nurseries grant you a second life. Literally. Instead of reducing the fixed glitched count when hit, the extra life is consumed.
	Factories grant a small wave of glitch fixing around the player, triggered by a first fix.
	Your typical glitch. Buildings sometimes turn into this and revert back once touched to fix.
	A sinister glitch. Does not contribute to your score when fixed, but instead creates deadly projectiles; glitch fixes need to be spent to recover from these hits or the player will ... glitch too! (This is game over.)

CHALLENGES & DESIGN

The game was developed for the lowrez jam, which restricts the drawing canvas -including the UI- to a 64x64 pixel grid. This creates challenges like using small fonts, and in general needing to fight for every pixel of space. I additionally took up the challenge with an unfamiliar to me engine and language; the DragonRuby GDK that is built on -you guessed it- Ruby. I was pleasantly surprised by how easy this was to use; core systems did not take more than a couple of days to implement.

The only real issue I had was that html5 deployment to play as  a web page. This ended up with a different framerate than the fixed 60fps of PCs (even after heavy optimization), and some manual time difference calculations were needed to make the game adapt to each platform.

Overall, I tried going for a very minimalistic UI that basically has two numbers; the elapsed time, and the current progress, that is, the number of fixed glitches . There are also a bunch of available effect symbols to appear near the progress. To simplfy controls without needing lengthy explanations (that do not fit in the resolution), everything is triggered by touching in-game objects, such as buildings to gain effects. Furthermore, I made sure to design levels where effects can be acquired at the beginning and it becomes immediately obvious what they do. Having short gameplay sessions (~a minute or less) lets players experiment with new. As a funny note,  the limit of 10 fixes was one of the very first design decisions and came about and was chosen due to the small UI; where making a larger score element to display numbers greater than 9 was visually displeasing.

Something that came up in the simulation (aka "stage") selection menu was that I needed as small a font as possible to display best completion times while still letting someone understand the numbers. Unfortunately, most fonts with 4 pixel character heights had some incomprehensible numbers. I was thus forced to create my own smol font visually optimized for understandable numbers when viewed out-of-context. In this small resolution, having camera shakes be convincing but not totally violent is also tricky. I ended up using a heuristic with exponential dampening of leftover vibrations,  but went through several variations of popular algorithmsbefore being satisfied. At the same time, permuting UI or in-game element sizes makes them too large, so I needed to resort to rotation, color changes, and particle emissions to make everything feel more responsive.

Resulution challenges aside, I wanted to have very simple screens and explicitly avoid having a tutorial for controls. For this reason, I basically incorporated the tutorial at each stage's beginning first moments of, where the player is prompted to use the mouse to move around, and -once they do- to fix 10 glitches. There is no delay to the actual gameplay and the prompts are non-intrusive, but they are still there as the occasional reminder.