You can find ALONE with K.I.T.T.I. here: http://www.ludumdare.com/compo/ludum-dare-22/?action=rate&uid=7263
What went wrong?
There was an unexpected family emergency that took up the better part of 10 hours. The emergency was unavoidable, and unforeseen. It was unfortunate, but I absolutely do not regret missing the 10 hours.
That said, 10 hours is a lot of time in a 48 hour competition. After sleeping for about 10 total hours and the family emergency, I was left with about 28 hours of total work time. Not bad, but it was not evenly distributed, requiring some energy drinks and long periods without sleep. If I had had the 10 hours, about 4 of it would have been for sleeping on Saturday night, and the rest would have been time spent on polish.
My first time sync when working on the project was going with the sprite.js library — it’s an amazing library, but I hit a wall when I realized the support for Tiled, which I was banking on, was not really all that complete. I was going to use melonJS from the outset, but then a couple of days before the competition, I decided sprite.js would offer a better experience. I had tested it out before the competition, and thought I could get everything to work based off of one of the libraries examples, but you know how that goes.
I ended up switching everything over to melonJS, which has faculties for reading Tiled maps directly. In the end, I think this was beneficial to the overall project, but it did eat up time. If I had just used melonJS from the beginning (like I had originally planned), I would have had at least another 3 hours of work time.
My second time sync was maps. I initially built a small test map, and iterated features over that map. The test map was an excellent strategy I think, but coming off of that I jumped into a really large map — bigger than I needed by several factors. I wasted a lot of time just trying to fill it in and make it playable. I eventually shrunk the map down, but I spent an awful lot of time with a map that I ended up throwing away.
I think if I would have just did a little more level design up front on a piece of paper, I could have avoided this time sync, and probably got several more hours to work on polish and other levels.
More experience with the Tiled editor would have helped to, but it’s pretty easy to pick up.
What went right?
Tools! I knew my tool chain fairly well — I’m certainly middle of the road competency wise, but I have been using gimp, vim, and javascript/html/css for quite a while and was at least comfortable with how they worked. I think this saved me a lot of heart ache.
The things I didn’t know well were sprite.js, melonJS, sfxr, Tiled and the music generating python script that GreaseMonkey posted about: http://www.ludumdare.com/compo/2011/12/13/if-you-find-it-hard-to-make-music-read-this/
However, I did practice using these tools before the competition, so I was comfortable using them, and they didn’t offer any surprises, or set backs (other than plain inexperience).
melonJS was fantastic to work with, and really easy to pick up with the great examples on their website.
Tiled is an amazing editor — combined with melonJS I think it really saved the project after the 10 hours I spent away from the competition.
And, of course sfxr really added a little something to the game. I’m no sound engineer, but having absolutely no sounds in a game is almost as maddening as eating at a restaurant without background music. sfxr is amazing. On top of that, it’s really easy to use!
And the Autotracker-Bu script from GreaseMonkey gave me something I thought I wouldn’t have. Perhaps the music isn’t perfect for the game, but it does add something.
Summary?
My experience? Positive. I’ll definitely be doing this again. Next time I’m going to hopefully avoid emergencies, pass on the energy drinks, get some nice evenly distributed sleep, and know exactly what my libraries/frameworks are capable of going into the competition.
Happy Hacking, and I’ll see you all in April! (and possibly before that for the MiniLD’s!!)
