What Went Right

1. Not giving up
Above all else, the most important thing that went right was the fact that I didn’t give up. After reading the theme I thought about it for a few hours before starting to work, and figured it would end up being a puzzle game (not much unlike a bunch of entries ended up).

However, puzzle games require lots of well-designed levels to be fun, and I was pretty sure that I couldn’t manage that in the time frame. There were several moments during the contest where I was asking myself why am I doing this, this is not going to be fun, but pressed onwards anyway, and the end result (after changing from pure puzzle game to something between a puzzle and a shmup) is pretty neat.

This is also the first time I’ve felt like continuing development on an LD entry right after the contest =).

2. Pseudo-random level generation
At first I had totally random levels, but they were, quite frankly, boring. I ended up writing a bunch of different level-generation algorithms (looped square, circles, grid, spiral), which then were fed some pseudorandom values with increasing number of partycles as the levels progressed. I think the first 25 levels or so are surprisingly good with this design.. after that we’re well into the bullet hell territory, and levels were turning pretty similar.

One thing that happened almost by accident is the fact that different levels require different kinds of gameplay - some are “almost” pure puzzles, some are “almost” pure shootemups, and of course the frantic bullet hell levels later on.

3. Time management
I managed to go to dog school, take the dogs out six times, take a shower, do some christmas shopping, sleep well over 8 hours in the middle, watch a movie, etc. and still submitted the final entry 9 hours before the contest end time.

4. Sfx(r)
One thing I’ve really improved upon in LD entries is the sound effects. Sure, I used the (amazing!) sfxr tool as basis this time, but also recorded (with a crappy webcam mic) and mixed some sounds. I used fractal music generator for the music again, which, while not good, is better than nothing.

5. Graphics
I think the graphics turned out pretty good this time, the little there is.

What Went Wrong

1. Bugs
A few bugs which I could have found (and possibly fixed) during the last 9 hours of the compo were left in. Widescreen monitor support, which I actually spent some time on, is broken. Midi music volume works differently than I expected, ending up changing the system midi volume(!), and third, some PCs, more-or-less-happily including my wife’s, are slow. As in 2fps slow.

I’ve solved the first and have been trying to figure the last for the past week or so. It appears that the rendering drops into software on my wife’s PC for some reason that I haven’t been able to figure out yet.

2. Tons of other stuff to do
Having this kind of contest in the middle of the holiday panic isn’t the best possible time =)

3. Tuning
Another thing where I could have used more time on; more level generation algorithms and/or hand-designed levels, slowing down the “normal” bullets from partycles, different kind of partycles that would lead to different kinds of puzzles.. etc. We’ll see if the game is popular, I might spend some time developing it further.

4. Code re-use, or lack thereof
I realized, while doing this contest entry, that I’ve probably written a bazillion font printers. Thus, I started working on a basecode set which includes most of the stuff I’m doing over and over and over again to be ready for the next LD.

5. Post-LD Exhaustion Syndrome
Returning to work and working on the multi-million line projects which move forward at glacial speeds after making a whole game in a weekend.. well, you probably know how that feels =)

This entry was posted on Friday, December 21st, 2007 at 3:56 am and is filed under LD10 - Chain Reaction. You can follow any responses to this entry through the RSS 2.0 feed. You can leave a response, or trackback from your own site.