I get an eye on the code there: Pushing Pixels and I do not understand its purpose.
Beside the speed differences, the returned image is the same as the passed one (here).
Am I correct ?
The examples are simple brute force copy routines… move the picture pixel by pixel from beachgraphics to copygraphics
the example itself has no realword application as there are much faster ways to copy an entire (or partial graphic)
BUT it does have (RGBSurface) a ton of uses…
scan a picture for a particular color and change it to another color
apply various photographic filters at a pixel level
My PaintDS program uses this to reduce color depth, change to various grayscales, calculate how to apply pattern based fill textures, rotate in 90deg increments, do gradient color fills, saturation, exposure etc. etc. etc.
So yes, it can be extremely useful, and if coded properly… very very fast.