#Ok - code architecture question - I can'
ruby orbit
You could probably do it similar to how actual screen blitting works - if you have a list of shapes in the order you get them, and have some loop that goes through them all and changes all of the pixels to that shape's color, the last one will win by default. You spend a bit of extra time drawing the same pixel multiple times, but ideally that would already be quite fast so it won't make a huge difference
glacial ermine
Ok so I guesss I failed to mention an important thing - shapes can be removed at any time in any order
so if we get shapes A, B, C, and D all overlapping on one pixel, in that order, and then C and D get removed, the pixel should end up being the color of B
onyx harbor
tbh just doing it and letting them overlap doesnt sound an issue at all, thats not really a thing you need to worry about anymore, even on modern phones
onyx harbor
ye
ruby orbit
Ok so I guesss I failed to mention an important thing - shapes can be _removed_ ...
If you maintain a list of all active shapes, and repaint them all in order after any change, that should handle that case too
glacial ermine
or I guess, the whole image
onyx harbor
overlap doesnt provide too much issues
glacial ermine
I guess the overdraw is probably not going to be a problem. I'll start with that and revisit this if it becomes a problem
ruby orbit
Along with some sort of clear flag that you set the whole image to before blitting shapes
onyx harbor
were in 2025, not 2005 anymore, sure optimizations are still nice, but i dont think optimizing this would be useful and just ends up in wasted time
glacial ermine
Alright thanks for this
onyx harbor
np but as digi said, dont forget to clear the flags and sort them correctly
glacial ermine
I think one minor optimization I can do without too much fuss is just to only draw the new shape when a shape is added
when a shape is removed, I have to redraw the whole canvas
onyx harbor
ye