Here we are – an hour from the submission deadline, and neither of us can believe we actually got our game finished. Judging by how the last 72 hours have gone, that wasn’t likely to happen.

This was our first game jam. We had discussed participating in a Ludum Dare before, but we only decided to go for it about two days before it began. We were so unprepared that we had no way to sync our work – Dropbox was full and we had little experience actually collaborating on GitHub. The one thing on our side was the theme choice, which ended up being the one I wanted. Once we started, everything… went down-hill.

The first problem was that we decided to make a first-person parkour game. In 72 hours. We chose a game with arguably the most time-consuming gameplay, buggiest systems and most complex art we could have possibly chosen. The second problem is that we decided to make a firs-person parkour game. In one room. Yeah, we thought that was smart at the time. We wanted to show off the main mechanic with which we utilised the theme of the jam – time travel. In Switch (the very uninspired name of our entry), the player can warp between two alternate… universes? Time periods? Whatever the explanation for warping, the two rooms differ in ways that allow the parkour to be used in interesting ways. While this idea seemed cool, it ended up being surprisingly hard to create interesting gameplay in such a restrictive environment, even when the rules for a ‘room’ are stretched a little.

Third problem – Unity. Some love it, some hate it. I swing back and forth depending in what time of day it is. In this case, it proved to be a struggle to use it. I’ve never built a scene this big before, so I don’t know what the performance should be like in the editor, but when just clicking on an object causes a RAM spike of approximately 10GB, anyone would worry. Being a UE4-centric developer means parts of Unity are counter-intuitive to my workflow, which makes level design a painfully slow process. I’m still not entirely sure if I’m just stupid or grouping objects in the scene is a very hard process. We had to use Unity however, as David (our programmer) has no experience with UE4.

That brings us to programming. Oh boy. controller.isGrounded will haunt me forever. Our parkour system ended up working, mostly. Yeah, you can climb through walls sometimes. Or climb up walls indefinitely by jumping in corners. Or the fact that you can’t jump unless you’re moving. But hey, it works… mostly. If you check out our GitHub repo you can see just how much ended up being commented out. One of our biggest problems was our movement being largely tied to framerate for a long time. When inputs only register properly if you press them on the ‘correct’ frame, it’s not a good feeling. For quite a while, finishing did not seem likely.

But there were good parts too. Hell, we finished. Making a game like this is hard any time, let alone in 72 hours. We did roughly what 2 years of development and polish produced in that Titanfall 2 level in less than a week. While our game is by no means the winning entry, we are both proud just to have made something that can be fun, even if it is short and stunted in potential. I remember the moment wall-running worked for the first time – I sat back in my chair and said, “Wow, this is actually happening. We made this work”.

stuff working


For a first attempt at game jams in general, we are both extremely pleased with our results. We’ve finished – I’m sitting at my desk, hoping the upload finishes before the deadline on my crappy 0.45Mb/s Australian upload speed while David goes back to his Hearthstone-ing. 3 days of stress is fading away rapidly, and I think it’s time for lunch. What an experience this has been.

  • Andrew Castillo (While False Studios)

Our entry here

The page

Our website

Leave a Reply

You must be logged in to post a comment.

[cache: storing page]