[quote=“opiop65,post:1819,topic:49634”]
No such thing.
A design document is a document… It doesn’t make your a professional or even good at what you’re doing.
Wanna be seen as a real game developer? Make lots of fun games. It’s that simple. :persecutioncomplex:
I was joking guys, see the winky face at the end? I don’t literally mean I’m a real game developer now.
Opiop, you ARE a real game developer.
On the subject of design documents, my worst work in my enterprise environment has always started with a fat design document, maybe even written by me. The more effort spent on the design document, the worse the end result.
My best work has been seat-of-pants stuff where I code in a frenzy of intuition, racing to finish the project before management notice and stops me.
Thank you!
I haven’t had a programming job, so I wouldn’t know how it all works but I know that I’m terrible at finishing projects without planning EVERYTHING first. In fact, my track record reflects that, I’ve finished 0 games (besides maybe Pong or small clones like that). Filling out this document sort of makes me realize what I need to do, and where to allocate my time and resources. I’m excited to get to work!
Shaving seconds off my Rubiks times. Learning bits and pieces of the Fridrich method has gotten my time down to 44 seconds! It’s funny, I’ve been keeping track of my times ever since I started trying to learn speed cubing and you can see a definite linear decrease which I’m expecting will flatten out shortly. Of course these aren’t my only solves, there’s hundreds of solves between these timing sessions.
I’ve been trying to find a way to balance my almost ADHD when it comes to project focus while still exercising my mind and fiddling with new projects.
Over the last few days I have:
- Made a simple serializer just for fun (by default only writes primitive arrays and primitives)
- Made a simple compressor that uses a dictionary lookup to index repeated words (good for basic compression)
- Wrote a standalone easy to use Ogg decoder that passes a RawPCM object with a buffer of raw PCM data
- Made a new basic game engine just for the fun of it (similar to Java2D in Graphics and Image objects) (this was before I started fiddling with new projects)
I will probably edit this if I do more this evening, but I just thought I should share my progress 
Finally got smooth following camera in Blox:
(The graphical glitch is copyright of the GIF recorder, not of my program).
Having started a module on control theory, I cannot see anything like that without thinking about transfer functions and feedback loops. Looks good.
finished another round on genetic-algorithms.
my last round was 2 years ago and i got stuck cos’ i was hiding the fitness function too much in the selection procedure - end up with a system which i could tell how to evolve but not where to - which was pointless :cranky: (debug images were looking better tho’)
became much simpler now, and it seems to work 
at least i could make it evolve a mathematical expression like (a + b * c) to create a desired number - based on an alphabet and bit-encoding. looks a bit boring (reading numbers on the console output) so i made a few color-mutations, which are not very useful but let me debug the code.
images read like : very left column contains the initial population. every column to the right is the selection/mutation/crossover of the previous generation. column on the very right should contain the desired color. each column shows the whole generation, sorted by hue+brightness which looks better than sorting by fitness
here is a simple almost-white input - evolving into blue :
averaged random seeds :
… red-ish input which is far away from the desired blue :
average :
and a random-ish color input evolving into dark red :
average :
inspired by http://www.ai-junkie.com/ga/intro/gat1.html
Hmmmm. It’s not evolution if you have a goal.
yea i thought so too. … maybe it is when you have a goal but no idea how to get there.
@ags1, basil_: In a world where being as blue as possible improves survival skills, everything will at some point look blue, so it kinda is, isn’t it?..
Had to make an implementation of the Heapsort algorithm for school and got a bit off topic after finishing it and playing with ways to get anything like an interesting result, which lead me to generate 1000 random numbers 0 to 1000 and then count the indices where [icode]a[i]==i+1[/icode]…
This number, although highly variable seems to almost always be between 10 and 60…
Edit: 38.3 in average.
Spent my day doing the problems on codingbat.com and writing a few helper classes around java’s XML DOM to make my life easier.
Did some basic 2D bloom. I think it brightens the game up quite a bit. Pun intended.
With:
http://puu.sh/ghvFx/2720d0c38f.jpg
Without:
http://puu.sh/ghvGO/94e37f6095.jpg
(1.0-0.001)^999 = 0.368063488
999 adjacent pairs, with each second item having a 1/1000 chance of matching the prior. Sorry, couldn’t hold back.
I don’t think that’s how bloom works. I can’t really see the difference except the brightness.
It should be making light brighter and kind of “overflowing” but not just making everything be bright.
I still have to adjust some of the variables, but if you zoom in you can see the glow. I added this (bloom * 1.75 * dot(final.xyz, lum)) to the final color with lum being vec3(0.2125, 0.7154, 0.0721) and bloom being a guassian blur of that particular pixel. Everything is done in one pass. Is that what bloom is supposed to be? I just messed around a bit and thought this pulled off the effect nicely enough.
try something like (bloom * 1.75 * dot(final.xyz * 2.0 - vec3(1.0), lum))