Linked Colours Post Mortem – Ludum Dare 30

Last week-end was Ludum Dare time again ! The 30th Ludum Dare game jam happend from the 23th to the 25th of this month and I was in !

Level1-Day2For those who don’t know, the Ludum Dare is a game Jam where you are challenged to create a game based on a theme disclosed at the last minute in 48h, alone and from scratch (the Compo) or in 72h with a team (the Jam). It’s technically a competition but having a playable game at the deadline is already the biggest win.

This time the theme was Connected Worlds. I took a little liberty in the theme as I set out to create a puzzle game where you have to connect all the squares of colours together to complete the level.

Play the game in your browser on here :

You can also. You better not use FireFox as the web player loads very slowly (Chrome is fine!), sorry for that inconvenience.

If you enjoyed it, please rate it ! Thanks !

You can also get the sources on GitHub.

What went right

Idea and development

As last time, I had the idea pretty quickly and stick to it. Thankfully I didn’t encountered big design issues and if I doubted that I would be able to create many levels, random generation came to my rescue, so I didn’t even needed to create levels manually other than the two tutorials.

In addition to Daneel, the framework for CraftStudio I created and which is the base of all of my projects, I was able to borrow quite some code from other projects of mine, so that I could focus on gameplay instead of reinventing the wheel once again. Having so much code already ready is what made me enter the Jam and not the Compo as the previous times in the first place.

But having an extra day prove very useful as even with a simple idea and lots of  ready-code I hadn’t much at the end of the second day in term of gameplay (but visuals where almost final).


Let’s say it with some pomposity: linked colours is the best-looking game I have ever created in my entire life, so far ! Look at this beauty ! I am not even an artist, or am I now !?

Well, not quite, but it’s what I am most happy about, the game looks really great, especially compared to my previous games (improvements, yeah !).

Random generation

At the end of the first day, I wanted to add a level editor so that anyone could have created levels. I have one almost working for another game but I ditched the idea even before beginning the second day, where I worked mostly on visuals. Instead I spend some time working on random generation, and it did go surprisingly well.

The generation is pretty simple : on a 5 by 5 grid, it picks a square, then picks a color, then finds another square nearby and sets a color the previous one can connect to. It repeats that until there is no more colorless squares on the grid. Then it picks the squares that are not connected to any one and change their color so that they can connect to a nearby square. Then it finds squares with 3 or more connections (3 is rare, 4 in really unlikely to happen, 5 or more never happens) and sets them as squares with a required number of connections.

And that’s it, and it produce pretty nice levels, with actual mixed difficulty. I just wished I had time to tweak it a little and have some variations in the grid.

Web Player

menuWeb players are very convenient, so I really wanted to produce a game that is playable that way. Unfortunately, CraftStudio‘s web player is sometimes a WIP and some things that works great in a desktop build just doesn’t in the web player, forcing me to find workaround when they exists. This was mostly what prevented me to release a web player build of Snale Driller, my entry for the last Ludum Dare.

But this time, I hadn’t had to add a single line of code to make it work in the web player ! It literally just worked… So, hooray !

Also, thanks to the web player, the game works great on Android (tested on a Motorola Moto G LTE running Android 4.4.4 and Chrome 36), even if the controls are tedious because the squares and links are tiny compared to my fingers.

What went wrong


Once again I didn’t worked that much since I only spent 22 hours (9+7+6) on the game.

And even if the game definitely looked nice at the end of day 2 I really didn’t had any gameplay but the two tutorials. It took me the “whole third day” (6 hours…) have proper gameplay and fixes.

Also time-related I hated the shortcuts I had to take with the scenes and code in order to finish what I wanted in time. But I guess that’s expected in jams to have not-optimal code (and general organisation).

For the same reasons, the game lacks feedbacks. I wanted to add more “juice” but I just added some sounds almost at the last minute. The inputs could have been improved too, like allowing to link by clicking and dragging the mouse instead of clicking every times.


random-endDefinitely more rights that wrongs this time, plus a successfully shipped game, nice looking and full of gameplay !

That’s a very nice addition to my portfolio, I am happy I was able to spend the time on this Ludum Dare !



Comments are closed