Better late than never. Here goes:
Graphics – This always seems to be hit or miss for me but I managed to find something simple and cute early on and mostly stuck with it. The style did evolve a bit over time so it’s not perfectly consistent but close enough to not be glaring.
Character – Somehow the combination of cute spherical animals, chaos and tiny planets came together well. I can’t say this was planned but I’m happy with the outcome.
Delayed decision making – I decided on the controls and genre for the game early on but didn’t really come to a decision about gameplay or goals until later. This meant I spent most of the available time on Saturday building the basic physics and movement of the game. This was a bit of a risk but it meant that the eventual goals and gameplay were tailored to what was achievable. I’m quite glad I didn’t commit to some of the earlier ideas I had that would have taken the underlying systems in a different direction as I don’t think they’d have turned out as well.
Using an existing engine – Choosing FlashPunk early on turned out well for me despite it being my first time using it (I’m generally a C++ developer, I’ve worked in as3 before but using my own entity system and supporting code). It was good not to have to maintain and fix things as I went and FlashPunk was kind enough to step out of the way when I needed to which made implementing the physics efficiently possible.
Ogmo Editor – Ogmo is great, very simple but a huge time saver when it came to building my levels and the XML output was dead easy to parse. The only downside (and this was not the fault of Ogmo itself) was FlashDevelop’s code completion feature throwing an exception and popping up a dialog every character I typed while editing the XML parsing code. For all I know this is fixed in a later build of Flash Develop.
Physics – By far the biggest problem pointed out in the feedback on my game was the glitches in the physics. I think it’s something I could have fixed easily but I weighed it up against the risk of destroying the chaotic nature of the game which added a lot of character. Maybe it was the wrong decision but it’s the one I went with. Unfortunately right after LD I didn’t have a lot of time to play with tweaking it for a post jam version and now it seems a little late. If I find the motivation I’ll have a shot at tweaking this but it will be a fine balance. When it comes to people I’ve physically witnessed playing the game it mostly caused laughter rather than outright frustration but this doesn’t mean there’s nothing to fix.
Time – I decided to join in late so I ended up doing LD on a weekend where I had existing plans. This probably chewed up about half of Saturday and 4 hours of Sunday.
Sound – This was a casualty of time, as mentioned above. I think it would have added a lot so it’s a shame.
In game control tutorials – In my first LD (#18) a common criticism was that the controls were given pre game and there was no way to return to them after the fact. I took this onboard and added the controls to an overlay that pops up around the player’s ship if they are idle for a while. Unfortunately as I found out watching people play the game post submission the common reaction to confusion over the controls is random button pressing so the overlay often won’t pop up. Regardless there was a lot less confusion over the controls this time so it’s definitely an improvement but can be refined further for next time.
In game signposting – I took this quite literally and added big billboards to the backdrops of the tutorial level to explain the goals. Unfortunately some inconsistency here lead to more confusion. As some parts of the early billboards looked like in game items but were non-interactive people thought that later billboards must also be non-interactive. This lead to people thinking that a merchant who creatures should’ve been delivered to that had a billboard pointed at it was for illustration only, and went off to find the ‘real one’. This was definitely a step in the right direction however and with some improvements to its consistency should work well.
Overall I’m much happier with this entry than any of my previous entries. It got more positive comments than I expected and I’m more pleased with it than I expected to be at the start of the weekend. I really enjoyed working on it and had fun playing with it myself. Hopefully I’ll be able to take on board the lessons learned from this to make a better game next time.
The game’s entry is here: http://www.ludumdare.com/compo/ludum-dare-23/?action=preview&uid=2573