Want to do fluid dynamics with fragment shaders. Should be hard but doable.
Presented by rdb
Presented by xmzhang1
Ratings (show detail)
Water Simulation 0.1
Well, my company is making me take python 101 training
So... right now I'm supposed to be learning how triple quote strings and upper() work and bored out of my mind. Guess I'll take the opportunity to blog about what I hope to accomplish.
- I'd love to make a machine learning game of some kind. Bit of a pipe dream in the timeframe we have, but maybe the theme will make an obvious case to fill out?
- Art is going to be hell, I'm so used to having teammates who are good at it. This year, I'm alone. Unless I drag my wife into it, and I don't know that she'd enjoy that considering she's already doing a seperate game art contest this month.
- I want to use my new (kinda expensive) music tools to make some awesome stuff. Or flail wildly at it, in particular all week and not have any time for code.
- I really want to lean hard into the theme, and make something that really beats a deceased equine organism with whatever connection I come up with. Some past occasions I've been super light-and-free with theming, and I want to avoid that this time.
Welp, gonna make everyone hate me
I think I've decided to use visual studio to develop my python code this year. The python plugins are starting to work pretty nicely, and having a GUI wrapper for different installs with info of what's pip-installed in them and able to spit out a requirements.txt automagically is nice.
Add that with shader autocomplete and it's actually a pretty satisfying experience to use Microsoft(TM) Visual Studio(TM) and Microsoft(TM) GitHub(TM) to work on pyglet stuff.
Now to see if I have any hope of coding fragment shader fluid dynamics next week.
Not gonna pull it off.
It's thursday and I'm still on basic simulation. This is gonna be a serious DNF.
My pressure system is great at filling new spaces appropriately. It's just terrible about emptying the spaces that flowed into them.
I've tried a "debt" based model that ends up running away to flood the world, and a "doom counter" approach that ends up static and unmoving. A number of hybrids that behave very boringly.
I've got one last thing I'm going to try before I stop programming for the week and switch to just playing games instead.
It may not look like it
But this took all week, and it looks very good in-program. There's no time for a game. All I've got is water simulation.
If anyone ever actually reads this, I developed some really useful re-usable code.
A back-buffer abstraction you can activate using
And a GLSL shader uniform parameter provider you can activate using
with UniformProvider(shader_obj, param=value,param2=value2, sampler=pygletTexture):
I made useful and cool things but no game :( Maybe inspiration will hit me sometime about how to make water pressure into a game, but I've been so focused on making it work, I haven't even begun to think of game ideas.