Yannick Remke

[Yannick Remke @YannickRtz]

The first screenshot shows the code to render the blue starfield with my library, the seconds screenshot shows the code for the same image in @ProcessingOrg . Results look the same. Third screenshot shows a draft for the new API that I haven't implemented yet.

[Yannick Remke @YannickRtz]

What I find curious: The order in which I intuitively put transformations is reverse to how Processing expects them. Probably comes from the declarative, assumption-loaded nature of my approach? Plus me having no idea about conventions.

[Yannick Remke @YannickRtz]

The first screenshot shows the code to render the blue starfield with my library, the seconds screenshot shows the code for the same image in @ProcessingOrg . Results look the same. Third screenshot shows a draft for the new API that I haven't implemented yet.

[Yannick Remke @YannickRtz]

What I like about my approach and want to keep:
- no classic for-loops
- resolution independent image description
- no pushMatrix, popMatrix
- stronger connection between masks, fillings and transformations

[Yannick Remke @YannickRtz]

Added support for transparent masks. An easy chage because of the simple (slow) way I'm rendering the image pixel by pixel.

[Yannick Remke @YannickRtz]

As always the code screenshot shows the complete code that is necessary to generate the picture.

[Yannick Remke @YannickRtz]

Added a generator for random numbers with fine control over seed values. Internally it's using Java std Random.

[Yannick Remke @YannickRtz]

Added the scale transformation :)

[Yannick Remke @YannickRtz]

Turned out there was a subtle bug in the alpha blending algorithm that caused some color values to exceed the 8bit range and overflow into the next channel.

[Yannick Remke @YannickRtz]

Some of the prettier bugs I encountered so far. The second was an error in the anti-aliasing code, the fourth had to do with rotations. The other ones I haven't solved yet, but it's something with transparency.

[Yannick Remke @YannickRtz]

Implemented simple anti-aliasing. It's about as slow as rendering at 4x the resolution and downsampling, which is basically what it's doing, but it looks nice. (Images are showing a detail)

[Yannick Remke @YannickRtz]

In the last image I used a sine function to simulate randomness. Guess it's time to add some random generator.

[Yannick Remke @YannickRtz]

Finally got around to adding a circle mask :)

[Yannick Remke @YannickRtz]

Had to restructure the code quite a bit for this to work. A good coding exercise, but I'm not sure if there's a real benefit to using this DSL over a state machine like syntax with simple loops like the ones implemented by OpenGL or @ProcessingOrg

[Yannick Remke @YannickRtz]

Added support for multiple variables in generators. The exact behavior can be customized using a second argument.

[Yannick Remke @YannickRtz]

Added support for different color representations: RGB 0-1, RGB 0-255, Hex-String and HSV

[Yannick Remke @YannickRtz]

Implemented proper alpha blending.

[Yannick Remke @YannickRtz]

Added support for procedural colors.

[Yannick Remke @YannickRtz]

Got rotations with variable pivot point working

[Yannick Remke @YannickRtz]

But I'm not planning on rendering only triangles in my library (...until I hit the next performance wall?)

[Yannick Remke @YannickRtz]

While trying to speed up rendering I realized two basic things about CG: Why there is a need for specialized graphics hardware and why in fast rendering most things are made up of triangles. It's just the shape that we have the best optimizations for.