What went right:
Overall I am really happy with how my 48 hours went, and with how my final game turned out. It was a great experience, which I will definitely be doing again next time.
Got off to a calm start. Ludum Dare start at 1pm Saturday my local time (New Zealand), and I think that worked out really well. I spent the first 3 hours having a nice casual lunch, and just started thinking, letting the theme absorb into my brain. I think this is the way to do it. Food helps me think, and since I went out, there was no way of rushing to the computer and starting before my idea was set in my own head.
Using Haxe, NME and the awe6 game framework. I am really happy with my technology choices. I have been using Haxe and NME for a while now for random things, never really use awe6 for anything other that simple hello world style games. I know everyone warns against it, and I knew it was a bit of a risk, but I made the decision to run awe6 anyway, as it seemed like the kind of framework that would not get in my way. I was lucky that I actually picked it up quite quickly, and was able to draw on the strengths of the framework to help me out.
XML based levels. Spent a little bit of time making a way to load levels as XML. This made making the levels really fast during the middle of the final night.
Simple game concept. Choosing a simple concept, and mechanic was a great decision. This meant I could concentration on making the game fun, and actually getting it finished.
What went average:
Needed more play testing. I really should have got a few more people to play the game. A lot of the things that did not turn out so well, could have easily been fixed it I got more people to play the game earlier on. That being said, I was tweaking the mechanics right up until the last minute, so not sure when I would have had the time to implement any feedback.
Maybe made it a bit too hard. I wanted to make a challenging game that would take full advantage of the mechanics. Unfortunately I think the difficulty curve was quite steep. There were a few levels that I probably should have made easier, or removed all together, but a number of people were actually able to play all the way through, so I think it is really up to if you “get” the mechanics or not.
Unusual jump mechanic. The jump mechanic that I was playing with, relied to much on knowing how it worked under the hood. I can now ace almost every level on the first try, but others seem to struggle. This was the main issue that the people who commented on my game post had. I should have worked a little more on making the jumping more intuitive. I have even started to tweak this for a post-compo build. A few people wanted to be able to move in mid-air. I purposely did not want to include this, as I thought it would change the game-play too much, and make it too easy.
Sound. Sound was average, as this is one aspect I am inexperienced in. I will definitely be trying to improve these skills for next time.
Not enough content. I thought I did quite a good job of keeping a reasonable scope, but the one thing that got away from me was the amount of levels/episodes I planned to create. My interpretation on the theme was that each level you would advance through the ages. Following the evolution of the wheel. You would start out jumping over prehistoric style stone wheels, then up to roman wooden wheels, then wagon wheels, and eventually to the future where everyone knows they use wheels made out of lasers. I only managed to get three different wheel types by the end, but I was not actually too worried about this. I think the point still came across. I will also add more wheel types and more levels in the post-compo builds.