Reddit
Islands in the Distance
So, a small collection of post-compo stuff here, timelapse, post mortem + bugfixes.
My timelapse is up at http://www.youtube.com/watch?v=NJ2UFsvPK4k if anyone’s interested. Looking back I spent a *lot* of time creating levels, but I kind of knew that already. It’s kind of fun to see them grow, though with the benefit of hindsight, taking more screenshots, and producing a ‘highlight reel’ after the event would lead to a more satisfying video.
What went right:
Had a clear design idea. I was lucky enough to start Saturday with a very clear idea of the game I wanted to achieve. It was (overly) ambitously large, and relied on being able to do a large amount of decent level design in a short space of time, but having such a clear vision of what I wanted kept me on task, and prevented me from wasting much time simply trying things out.
In-engine editing. I’ve been converted to the idea of doing level design in game ever since playing around with Cube god knows how long ago. So early on I wrote a system to support editing of levels without leaving the game (I didn’t even have an edit mode to flick into, editing happened literally at the same time as playing). There’s simple no way I’d have been able to design as many screens as I did if I’d had to keep flicking between level editor and game (or, worse, text editor and game).
Simple mechanics that interacted nicely. The 3 core mechanics (springy things, ice, and rocks) were chosen mostly on a “what can I implement quickly” basis, but I was very happy with how well they ended up slotting together.
What went wrong:
Too damn big. Whilst having a clear idea was great the specific idea I had committed me to making waaay more content than was comfortable within the timescale. There are 40 screens in the final game, and nearly as many puzzles, and this is *after* I’d scaled back my original vision significantly. I took a huge chunk of total time, probably approaching 8 hours on pure level design. I ended up having to cut other things I wanted to put in (extra animation, localised fauna etc.) as a result. Having said all that it was really cool to realize something on that scale.
My music is dull, repetitive and quickly annoying. Just a combination of of it being a generally a weak area for me, and having very little time to burn on it (1hr fairly late into Sunday, when I was completely knackered). I came very close to not bothering including music at all though, so I guess overall I’m glad I managed to put something in. More practice required before next time though.
Less testing time than it could have done with. There ended up being some significant and unpleasant glitches (infinite-loops, an inescapable area etc.) that I just didn’t have enough time left to find. Ties pretty heavily into the “too damn big” point, with that much content there’s going to be problem areas, and it’s going to take hours to find them.
Not taking Monday off work. Simple, but regrettable, trying to do a full days work after a 36hr game development marathon was not fun!.
So, all in all, I shot pretty high, way too high in fact, so it feels pretty great that I managed to realize my original idea fairly convincingly.
The glitches + bugs have been annoying me a lot, so I’ve been working on a fixed up version, available to play at http://jwhiting.nfshost.com/ld/17/islands_better.html. So far it has fixes for all the specific glitches people have reported (thanks for the info!), as well as a few I found myself. Also I’ve put in code so that if anyone does run into an icky infinite loop situation it should at least recover from it fairly gracefully. I also plan to add some sort of progress saving system, so that it doesn’t have to be completed in one session.
Anyway, that post was pretty epic all told, congratulations if you made it to the end 
Now I’ve just got a lot of games to play 