A PostMortem for Windows Cleaning Simulator

Posted by (twitter: @ranska_)
December 12th, 2014 6:59 am

There 3 of us in the team

Pierre-Alexandre : dev
Léo : 2d/3d artist
Me, Simon : Game Design and Musics.

It was our first Ludum Dare and we liked it, we love the game we made but everything wasn’t smooth and simple.


We mostly prepared our workflow. A Trello to assign tasks and a bitbucket (git) to centralize the work and make the deploy easy.

We also tested the deploy, because we make html5 games, and because web is sometimes a pain in the ass when you tried to deploy app, so we had to test that before the LD was launched.

In retrospect, we were prepared and all this did help us to get through the LD. But not enought. We didn’t talk about methodes : how do we build a level in a game, how do we decide limits of polish in relation with half-days updates. Well we didn’t talk about all the process stuffs.

Day 1

Up at 3 am, the theme was just reveal in France, we brainstormed until 7 am, wrote about 20 ideas (3/4 okay for a game). Then we decided the first deadline for the evening : we have the main screen (the building) and it can be clean. Which was too much. Too much dev : level + cleaning anime (no implemented finally) + windows that can be clean, and too much graphics assets (because we did all 2d pixel-art with Blender).

And more the day was running less graphics and dev progress.

At the time we didn’t really understand why.

Gone to bed near midnight after a few games of Race For The Galaxy (nice card game by the way).

Day 2

At this point I have to say that I am working with Pierre-Alexendre on a bigger game so we are used to work together, but this LD was our first time working with Léo.

Pretty much the same morning but we’re up at 9h (too much sleeping was needed).

At noon we realized that the dev was not in a good way because we have not enought graphics assets to move forward. Two reasons for that :

– Léo didn’t know how to produce an asset useful that match my mock-up of the level (we used tiled and .json file for processing the level).

– The beauty of the graphics needed a work that take time, time that delayed all the dev.

Those things could be managed early if we knewn eatch others works habits and needs. Our inconscient way out was to go on Léo’s rythm and keep our fingers crossed.

Luckely everything was peachy during the afternoon, we switched from tilebase to heavy sprites, rewrote some parts of the code and by the end of the day the dialog box (text) system was ready. Musics were ready and all the game design and texts (from the day before).

Day 3

Up at 9, we started working near 10. The point of my personnal day was to get a playable and endable (is that a word ?) version of the game. So cleaning (without anime) and 2 endings (with actions link). Nothing fancy.

But this is the day where we payed the delayed work from day 1. We made the 0.8 version, the game was like he is right now without animes and some lower graphics assets.

At 6 pm, we’d to go to present the Ludum Dare and our game to a fablab near by (The Open Factory).

From 9 pm to 2 am, we implemented offices’s animations and made some polish on the front building.

And the game was over.


Planning (even when we failed to achieve our goals) and cutting features (Joseph’s animes, sound effects and a few dialogs cases) were things that saved us in this Ludum Dare. Most of our problems was not the engine (Phaser) or using Blender to do 2d pixels arts. But our inner communication as a team and our different understanding of the goals that we’re aimming for (the goals in the production process not on the meaning of the game).

Thanks for this Ludum Dare we really have great time making this game and we learn a lot.

Leave a Reply

You must be logged in to post a comment.

[cache: storing page]