Post mortem

About a year or ago I watched a few live streams and got curious. I tried one LD on my own without creating an account, knowing I had no chance. It ended badly as expected but I had quite a lot of fun and was surprised at how far you can get when focusing so much like that. This is my first official Ludum Dare, and you can bet it’s not my last :-)

What went right:

• I think I had a nice concept that fit both themes.
• My usual #1 enemy is scope creep, so even before knowing about the theme I had decided that I’d have the smallest scope possible. My top priority was to finish, and I did! Next time I’ll do better :-)
• I did a few test scenes where I threw in and isolated some of the game objects just to test a script or feature. It ended up pretty useful when I got a regression bug the next day. I just went back to those test scenes and debugged it in a pinch. Also, once I got the movement script of the enzymes working correctly in the test scene I just loaded the main scene and it all just worked. It was magical :-)
• I used a few Gizmos and a lot of Empty Game Objects (I call them Markers) instead of hard-coding positions and rotations. It helped a lot to quickly make 4-5 layouts.

What went wrong:

I never made it that far before so I had lots of little surprises

• I didn’t know a thing about aspect ratio so I wrote the whole game in “free aspect”, in a totally impossible ratio… Then got scared learning the hard way how much of a pain this thing is…
• I lost too many prototypes due to bad code structure so I spent countless hours fiddling with best practices, MVC and MVVM and messages and signals and events and delegates and other ways to structure the core of my games, I watched hours and hours of tutorials and live streams, and I felt like I finally had this nailed down. Well, it certainly helped but I still don’t think I nailed it… Back to the drawing board :-)
• I had never realized that building a system is just one part of it. Tweaking and testing and tweaking and testing some more takes way more time that I though! And changing one little thing like the movement speed can break a lot of things (I also had to adjust the rotation speed because stuff was clipping, for example).
• Randomizing stuff is nice but it’s not good for gameplay. I fixed that in my post-deadline version but in my entry, the type of nucleobases showing up is totally random so you quite often end up with having to do a bad move even though you know it’s wrong.
• I thought “1h is way more than enough to submit” so I let my guard down and took a break… I finished submitting less than 5 min before the deadline! I got really scared when I saw the time it took to build (especially the WebGL version!), package, compress and upload all that stuff.

That’s it! Please go take a look at my game and tell me what you think! http://ludumdare.com/compo/ludum-dare-34/?action=preview&uid=31782

See you next time!