That Damned RGB Triangle: Where Every Graphics Career Begins
Description
A four-panel meme using the '13 Reasons Why' party scene format featuring a young woman with long curly hair holding a red Solo cup. In the first panel she says 'So you see, that's where the trouble began.' The second panel shows the iconic RGB-interpolated triangle on a black background - the classic 'hello world' of graphics programming, with red, green, and blue vertices blending across the surface. The bottom panels show her saying 'That TRIANGLE' and 'That damned TRIANGLE.' The joke is that rendering this single vertex-colored triangle is the gateway drug into the deep, painful world of graphics programming - OpenGL/Vulkan boilerplate, shaders, and GPU pipelines - and everything goes downhill from there
Comments
30Comment deleted
1,200 lines of Vulkan boilerplate later, you too can render the triangle that ruined her life
Three vertices in, five years later you have a custom allocator and no game.
Instead of opening Pandora's box, they OpenGL Comment deleted
speedrunning Raylib Comment deleted
a black-red triangle? Comment deleted
Should have used Comment deleted
No color spaces in this space please. Comment deleted
YEAH WHITES ONLY Comment deleted
[This silly sarcastic post brought to you by a brown person] Comment deleted
oh hey a fellow system. hiya ^^ Comment deleted
Meowdy! ^^ Comment deleted
Damn Comment deleted
This is an English only chat, please adhere to the rules Comment deleted
I missed the chat 🙈 Comment deleted
I was wondering, where the f my photo went 😅 Comment deleted
You're not getting the joke Comment deleted
Please explain Comment deleted
The triangle in the meme is the first thing that's rendered on the screen when one's learning graphics programming, whether it's OpenGL or something else. Graphics programming seems easy at first but soon enough turns out to be a massive rabbit hole with lots of headache guaranteed. So yeah, all the trouble begins with that damned triangle. Comment deleted
Ohh, I see, thanks Comment deleted
Honestly found it easier than working in engines, aside from 3D stuff Comment deleted
I just hated every second of OpenGL classes in uni. I had to choose between Chaos Theory and "introduction into computer graphics". Should have picked the damn math... Comment deleted
"The damn math" Seems like you either don’t love triangles enough or math, or both 🤓 Comment deleted
triangles, as in three angles, Euler angles Comment deleted
Hey! To meme this meme was about ONE MORE step forward Graphics and games are good for joke starter, but without billions of investments in GPU we would never have all those sophisticated LLMs 🌚 Comment deleted
Hate to explain jokes but gonna intervene when someone may think that this channel finally posted something not about AI 🌚🌚🌚 Comment deleted
Not only is xy diagram the worst thing to judge color gamut (mostly ergonomic improvement on the original rg diagram), but the whole idea of representing volumetric information in a planar view, even with more advanced u'v' or a*b*. 🧐 Comment deleted
Its a middle slice where most colors are perceived, 50% lightness (in hsl terms or smth.) or smth. Because with more light it all goes to white, colors are lost. The same with less light - all goes to black, also no colors. Talking about other shortcomings - well , thats our eyes, dread it, run from it, destiny still arrives. Comment deleted
Absolutely not! Color gamut of a device has complex shape in a perceptually uniform space like L*a*b*, with different luminosity levels for each primary stimulus (RGB, CMYK or whatever), so there is no single "slice" where the chromatic saturation is at maximum. Color perception is not static and depends on multiple factors, both external and internal to the observer, so any given stimulus can be perceived differently according to many conditions. Comment deleted
I was talking about classic cie 1931 that describes human eye perception not device gamut Comment deleted
It was the first attempt to objectively measure color perception, in the times when there was no possibility to inspect the spectral characteristics of retina directly, as well as the output signals it produces. Lack of computing power led to simple formulae. However, even if that was feasible at the time, color perception is much more complex than that, because it goes all the way to the cerebrum: the basic output that the eye produces (which is not trivial by itself) is further "adjusted" by the brain. Even the most advanced models developed currently, like CIE L*a*b* 2000 (1976 space + ∆E:2000 difference formula) and CIE CAM 2016 cannot provide flawless result in the general case. But using 1931 models nowadays other than for purely technical low-level data representation for specific uses is totally lame. Please, refrain from it and do not spread ignorance. 🙏 Comment deleted