What I did today

Today I finalized some of the specifics for my Behavior Routine AI system. While this definitely wasn’t the most challenging thing I’ve done so far, it has definitely been the most insanely frustrating. It involved the realization that a few of the systems I created a few years ago were absolute tosh and nigh unusable and needed to be tossed out for something better.

But the end result is pretty okay and I feel verily satisfied.

I’ve also been putting up a bunch of free music on my Soundcloud for people to use during Ludum Dare if they so desire. Not much cohesion between the different tracks, but an… okay amount of variety. It’s fun, and making music for #RFLEX has helped me broaden my horizon (mentally. I didn’t believe that music was something I could do since I’d never done it before.) as far as what I’m capable of producing. These past few days have been tons of fun, and I am really looking forward to the future.

tl;dr: Free music, AI is nightmarish, gettin’ better at moosik

Damn , impressive . Biology really isnt my strong point.

Started my new job at a local bike shop called the bike smith, woot.

Good bunch of guys, good laugh, trade prices on stuff :p.

Blogged. Because I’m sad.

http://opticsandalgorithms.blogspot.nl/

Solved the shadow problems in engine ;D

Now I have horrible performance again though :’(

Enemy redesign trying to make the game flow together:
http://zippy.gfycat.com/CloudyInfamousAmphibian.gif

Added enemy walk and idle animations, as well as updated the enemy model. Got new textures for the sand, walls, etc.

Made the enemy stop before it intersects the collision sphere to prevent the enemy from pushing the player around. Made the enemy play the idle animation once it has reached that point before collision.

Reworked lighting a bit, and touched up the software renderer to improve speed (slightly).

I upgraded my directional light shader and cascade handling code to handle a variable number of cascades as well as avoiding all branches in the shader. Before, all directional lights used 4 cascades. Now, by changing a simple variable, I can configure a specific directional light to have 1 to 6 cascades (point lights need 6 shadow maps, so up to 6 can fit in the same shadow map). Shadows for the whole scene is now possible!

Next up: Frustum fitting to fit each cascade to the depth range it’s supposed to cover instead of wasting 75% of each cascade’s area.

Today’s simulation because I was bored: [extremely] crude n-body-like:

BurntPizza that looks really cool!

Over the past couple days I’ve redesigned my personal website (again), started to set up my site on my uncle’s VPS, wrote the first draft of my resume and almost quit my dishwashing job because it’s just bad. Just have to stick it out until my internship… :frowning:

Learned a bunch more about compilers and how they turn nasty code into nice, optimized ASM. Very beneficial for my RE skills!

Accidentally hit a curb going ~5mph and had to run to the nearby hardware store to get a saw so I could take off a black plastic piece hanging off my bumper… it’s not a good feeling to hack apart your car. Luckily it’s ok though, nothing is damaged!

Sketched a jungle background for LD:

Finally after touching it in Inkscape:

Woohya! Art is done. Polishing next.

Figured out how to change mouse sensitivity from within the game yesterday. It required grabbing/hiding the native cursor and replacing it with my own, so now I have a hilariously misleading placeholder! But you can control the cursor with a controller now if you want. :open_mouth:

No offence, but that’s bad design. Majority of people change sensitivity for cross-hairs and head-movement in game. But people don’t do it so their UI cursor is faster for a specific game. (Other than the minority of people downsampling their displays to a higher resolution, but the graphics drivers that handle it usually tweak the mouse sensitivity auto-magically)

Look at battlefield and minecraft and how they handle their mouse sensitivity, they do it right.

Found out that the “small” game I’m building is too big for me to wanna finish, smh. Guess I’ll be back at it tomorrow.

I finally got to the point where I could release a minimal build of my program XShot, which I am using to learn more about Java design patterns and software. Jumping into real game dev with Ghostr, using weird physics and Box2D was a bit much, so I decided to work on this project instead. But after this is done in a few weeks, I’ll give true game development another go with a simpler game.

Submitted the first version of my bachelor thesis on the order-independent transparency algorithm I came up with.

EDIT: Reworded because words are hard.

I finally finished an LD game attack/defense. I did it with HTML5, and I think I’ve fallen in love with Javascript.

F

[sup]:P[/sup]