Galapagos Post Mortem
Here’s a timelapse link for my LD17 entry. Sorry no music on it.
What went wrong? What worked? Full post-mortem writeup after the bump… including a SNEAK PEEK screenshot of the progress on the game!
The Plan My plan (which is pretty much my plan every time) is to get an idea and a minimal prototype working friday night, then build out the engine on saturday, leaving sunday for content. This plan failed. I didn’t even have a solid idea of what the game would be until halfway through Saturday. I really liked the theme, in fact it was a bit too helpful, I had a lot of trouble narrowing down my ideas. What went Right- Art Style — One of my goals for the contest was to make my game have a distinctive art style, and that really worked out. In addition to making it stand out visually, I also chose a style that made the art a bit faster to create. Even so, I didn’t have a lot of time to make art but I think it looked as if I did.
- OGMO — Tried out the OGMO level editor for this compo. It’s great. Get’s a bit slow on large levels but if you keep it small it’s a great tool. The flexible ‘value’ system saved me a lot of coding.
- Music — Around midday Sunday I was faced with the choice of add some music and certainly not have time to finish the game, or don’t add music and still probably not finish. One of my goals this time was to do music, so I went for it. I think that not only did it add a lot to the game, it also helped mentally recharge me and allowed me to be even more productive Sunday afternoon.
- Tweakval — I used my tweakval code again, and it saved me hours of work, especially in tweaking the projection. Getting the “simple” ortho projection right was surprisingly tricky.
- Idea — I took way too long to decide on an idea. Even by Saturday night I had a good sense of the look/engine I wanted but only a vague idea of the gameplay. “roguelike” isn’t enough detail of a design. I should have sketched something out on paper on friday, and stuck to it.
- Not doing gameplay first. The key to success is first do player interaction, second do enemies, then work on map and world building. That way, if you run out of time, you still have something playable. But somehow I always seem to mess this up, probably because I have a lot more practice writing the map/world kind of code than gameplay code. It’s a vicious cycle. Next time, right?
- Blender — This didn’t really go “wrong” because in the end I was able to build my models but it took me like an hour to build a half-sphere (most of that was because I had specific ideas about the UV layout). Really I just need more practice, but it was a time sink.
- Basecode — I don’t really have a formal basecode, just a pile of snippets and functions from past LD’s and stuff. I spent a bit of time debugging some stupid graphics bugs and redoing basic stuff. Hopefully I’ll have time to clean up and simplify this before the next compo.
The Future of “Crater Galapagos”
Here’s a sneak peek of the current post-compo game. It will be a while before I release anything though:
