Computational expense question (2)

Processing Coding Questions
1

I have a very simple and small image, easily generatable with code alone (it’s essentially just 4 differently colored squares and the image alone is 32x32 px), and I wish to know what is more expensive:

  1. using textures along with textureWrap(REPEAT); to tile a part of the canvas with this image, or
  2. generate the image with code and use for loops to do the tiling instead.

To give you an idea, I’m trying to create a drawing canvas, and I need a background texture, similar to that of transparent images, with the white and gray checkerboard square pattern. The texture obviously mustn’t stretch, it should appear like a static background, and as you adjust the size of the canvas, you “mask” parts of it away. I did think of creating an image with the same size as my screen and displaying that first and then masking it, but I want my program to be compatible with multiple screen sizes, which means no set image size, but rather a tiling approach.

I also want to ask one more thing in advance. Another element I will be adding to this program will be a resizable square-shaped window, on which I’m planning to draw some sort of 3D canvas (the ultimate goal will be to have a 3D rendering of a sphere on one side of the screen, and a 2D map of the surface of that sphere on the other), and I need to know what the best way of going about this is. I’m not sure whether PGraphics allows for a resizable canvas, so I’m guessing the only other way is to use PGraphics in conjunction with a PShape or vertex-related functions, which I will texture with a PImage rendering of this PGraphics canvas.

As in:

create a PGraphics called pg
draw on pg
beginShape();
texture(pg);
draw the 4 vertices
endShape();

Will this work? The fact that vertex() allows for automatic mapping of textures onto the shape certainly helps a lot. But more importantly, is there any other way?

2

i have not experimented on it, but I think the final result will be something that you display with

image()

in the size of the screen anyway.

I think that in any case it can be a single event - producing the background image once and display it as the first item instead of using background().

since it is probably a single method you call now and then, it could even be a

thread()

, as to not stop your running code.

you could also try to fill the pixels themselves, but in any case you don’t need a PShape at all.
creating a new PGraphics with the current (new) display size, filling the background white and drawing black rects, should probably do the trick imho.