Reddit
Pocket Strife: GIT timelapse
Session 1 [ Fri Apr 20 20:56:25 -> Sat Apr 21 02:22:08 ] 5.5 hours
Two hours in:
Figure 1. iso-mine-craft shows it’s sexy side.
At this point I felt pretty proud of myself. I’m two hours in, and have something that #ludiumdare and #love actually like the look of. If there’s anything I learned from minecraft, is that a lot of low res crap makes one big hig-res-low-res thing that looks good. Of course, I’m totally bombing the framerate here.
Five hours in:
Figure 2. I’m starting to realize that this was a bad idea
For the next three hours, I spend the night trying to figure out how the hell to a) make the performance smooth (e.g. only draw the tiles on screen, don’t just test if they’re on the screen) and the math to make the mouse work. I’m getting at best 46 fps (where it should be closer to 1000 in this case). I’m getting discouraged, and decide to go to bed, as I have a friend from California coming in, and a band to try out for that night. I spend the night chilling, playing guitar and smoking shisha with my girlfriend. Obviously I’m not taking this too seriously, but I did plan on using Sunday to finish the LD and make a “game” out of what I have.
Session 2 [ Sun Apr 22 ~13:30:00 -> Sun Apr 22 20:24:06 ] 7 hours
Eight hours in (5.5 left):
Figure 3. Now we’re getting somewhere!
I came back and slept in until 1:30ish on sunday, and started at it again. I decided to use all the shortcut tricks I knew to get the framerate back up, and used some seriously dirty hacks to get the mouse to work (had to get the location during the draw phase, and then hide the mouse to avoid suspicion!). I’m not proud of them, but it beats spending hours trying to figure out the math myself (sorry internet, you weren’t very helpful). At this point I started bashing away at the game, and things moved very quickly. I had four building types, I integrated the health bar, and you couldn’t place buildings on tiles that were too blue or white. Later, because this wasn’t dynamic enough, I wasn’t able to include the cool maps and such, like lava planet and the moon.
Nine hours in (4.5 left):
Hard coding only four items really came back and bit me in the ass, but I knew what I wanted for game play, so it was not worth the effort making the system dynamic. At this point, I have added the territory flag, but I still haven’t gotten the AI in yet.
Twelve hours in (1.5 left):
Now the AI is in, and working at full stupid. Using a bit of math, and tweaking it until the computer made what I wanted it to make, life was born. The computer is faster than crap, and doesn’t really think much, but it’s there, it’s pretty, and it’s hard to beat. I like that. At this point I start integrating the lovemenu library that I have.
Let’s not forget the kitten challenge! I spend the remainder of the time testing the game, and fixing small bugs.
Thirteen hours in (0.5 left):
The game is done, and it’s time to start packaging for distribution. I take pretty screenshots, and shove them into my repo. I run my build scripts and upload the final product up to mediafire. I fill out the information on ludumdare and leave to have burritos at Moe’s. I was unimpressed by my framerate and graphics, sad that I had to include the mouse hack, but still rather proud of my product.
Total: 13.5 hours
Tags: timelapse