Hello all,

tl;dr at the bottom

Brief Background: I REALLY wanted this to be the compo/jam that made me complete my first game.  In August, I attempted LD27 (10 Seconds) with very little success.  Very little preparation, bad timing, and the time I had during the jam I used it to learn SDL 2.0 along with dividing the C++ into classes/files instead of 1 giant file. I had gone through SDL 1.2 tutorials shortly before and all I made was an unfinished glorified pong game that had flaws with detecting collision when the ball was coming at a perfect 45 degree angle at the paddle and the paddle could never detect it (still not sure how it would accomplish this… maybe a “look ahead” of some sort?  I wish I spent more time from that jam until this like I had wanted, but I think I was let down among other things.  I have also been upset at myself for not getting back into gamedev sooner.

This past week prior to the jam, I knew I had to prepare and get back into it if I wanted any chance this LD.  I started a beginner RPG tutorial and learned some good things. Some of the Object Oriented stuff I tried on my own last August was the same but better so I thought that was cool.  I  hadn’t really done C++ since ~2007 in school.  I didn’t finish the tutorial until 7 hours after the compo started, which I was OK with but preferred to be done with it prior. First mistake.

Saturday comes, and I wanted to wake up around 10 or 11am (6-7 hours of sleep), but instead it was 2:30PM! I should have set an alarm.  Long work weeks with little sleep caught up to me like usual: Mistake #2. 18.5 hours had gone by since LD started and I’ve yet to start on the actual game. I had a few ideas when the theme started but I was focusing on finishing the tutorial/engine.  I started with my OK base code and started to add my own code into it.  I added a FPS cap of 60 and it fluctuated between 35-55, terrible.  Then I tried to add a cap of 30 and it was around 24-27.  I toyed with a few things and it wasn’t working well and I figured I didn’t want to make a hack&slash because almost 24 hours passed and it wouldn’t be enough time and b, I’ve mucked with the code enough I wanted to start from a clean base.

Hours 24-41: I backed up a little bit and due to time remaining and lack of knowledge, I just wanted to get a nice simplistic game and jotted down a design. I added to existing engine from RPG tutorial without worrying about framerate issues, deleted and backtracked the camera engine portion to get back to static with moving characters, drew some new assets and displayed them on the screen. I then fell asleep for 11 hours yet again! Sometime the night prior, and the time from hours 41-48 I was then able to draw my player, him on a paddle/platform, a bullet in his mouth where he shoots from, and an enemy on a paddle with a bullet on the opposite side. But collision detection was failing miserably and there goes the compo. I was also able to get my first main menu going so that the user has to press SPACEBAR to start the game. I was pretty stoked about that (little victories and firsts).

Due to collision issues and mess of engine and game code, I decided it was time to revert back to my original code from August of SDL 2.0 mixed with SDL 1.2 and retrofit/combine the good from that with the good from my current RPG engine turned into static background, basically starting from the ground up.  Took Monday off of work and worked on the game all day to try to get it working until about 750pm (jam ends at 8pm) where I got so close but barely ran out of time. I got some of it working including collision and could possibly finish it in a few hours without music or sound effects and I will definitely continue working on it until I complete it. I didn’t want to submit a game where a couple sprites just bounce around the screen.

Lessons learned and tl:dr

What went right:

1. I prepared a little earlier than before which helped me actually work on the game during the compo
2. Hours 0-2 of the compo, I actually went to the gym to do my brainstorming of the theme while working out, which helped me.
3. Saved versions of the project during critical times.
4. Taking a step back at the right time to regroup/replan and refocus.
5. I did 1 status in the middle when I got my menu to work, what needed to be worked on next and potential add-ons, which helped prioritize.
6. Clean workspace and good retro MIDI tunes playlist picked out prior to the jam helped.

Things to improve:

1. Preperation. Although I prepared better than before, it wasn’t enough.  The 5 hours spent finishing learning at the beginning of the compo should have been spent working on the game.
2. Also, axiom keyboard not working and should have gotten that resolved prior
3. Work on pixel art prior.
4. Do a warmup of a game with just squares even to make sure basic engine is working.
5. Either learn Unity2D to not deal with engine aspects during the jam or have a solid engine done already.
6. Manage sleep better the week leading into it, which would have earned another ~10 hours total of compo/jam time.
7. Overall, complete some small games first to get out of the way and get a better feel of everything.
8. If above is improved, I can actually get people to playtest the game prior to submission time to work on bugs.
9. If above is improved, I can also have time to add next phases into the game (noted in my 1 status post) which would have been fun.
10. If above is improved, more status posts.

This has been a pretty good learning experience, fairly better than the last compo and can’t wait for the next one.  I only get TWO failed Ludum Dare attempts and I foresee nothing but completed Ludum Dare attempts in the future.

My menu prior to starting from the ground up.

Tags: LD28, post-mortem, postmortem