Thank you so much, @noel. The Floodfill looks great!
To fix the import issue on openprocessing, I think we could replace java.awt.Point with PVector. Could you try that? We might also need to / want to change the linkedlist / deque into ArrayList, I’m not sure.
That’s already done. @GoToLoop made an excellent class. but unfortunately, it’s still not running on OpenProcessing. See here.
Wow, this looks great fun. I am looking at all the unfinished topics looking for one to tackle. The Rosetta project shows the aesthetic beauty of code ipso facto. Sometimes the code is more beautiful than the output and too often hidden.
I would like to play too! 
from collections import deque
tolerance = 15
allowed = False
def setup():
global img, fill_color
size(600, 400)
img = loadImage("image.png")
fill_color = color(250, 0, 0)
fill(0, 0, 100)
image(img, 0, 0, width, height)
textSize(18)
text("Tolerance = {} (Use mouse wheel to change)".format(tolerance),
100, height - 30)
text("Right click to reset", 100, height - 10)
def draw():
global allowed
if allowed:
image(img, 0, 0, width, height)
text("Tolerance = {} (Use mouse wheel to change)".format(tolerance), 100, height - 30)
text("Right click to reset", 100, height - 10)
allowed = False
def mousePressed():
global img, allowed
img.loadPixels()
flood(mouseX, mouseY)
img.updatePixels()
allowed = True
if mouseButton == RIGHT:
img = loadImage("image.png")
def mouseWheel(event):
global allowed, tolerance
e = event.getCount()
tolerance += 2 * e
if tolerance > 256:
tolerance = 256
if tolerance < 0:
tolerance = 0
allowed = True
def flood(x, y):
target_color = img.pixels[pixel_position(mouseX, mouseY)]
queue = deque()
queue.append((x, y))
while len(queue) > 0:
p_x, p_y = queue.popleft()
if (check(p_x, p_y, target_color)):
queue.append((p_x, p_y - 1))
queue.append((p_x, p_y + 1))
queue.append((p_x - 1, p_y))
queue.append((p_x + 1, p_y))
def pixel_position(x, y):
return x + (y * img.width)
def check(x, y, target_color):
if x < 0 or y < 0 or y >= img.height or x >= img.width:
return False
pp = img.pixels[pixel_position(x, y)]
test_tolerance = (abs(green(target_color) - green(pp)) < tolerance
and abs(red(target_color) - red(pp)) < tolerance
and abs(blue(target_color) - blue(pp)) < tolerance)
if not test_tolerance:
return False
img.pixels[pixel_position(x, y)] = fill_color
return True
@noel @villares me too. See on rosetta code, I used suggested image, and this is the result:-
I’ve added a link to play the Bitmap/Floodfill task here. Unfortunately, if the Device Pixel Aspect Ratio of the monitor is not 1.00, the sketch will probably not run properly. This happens for instance when you chose a different text percentage (like 125%) in your “windows appearance” settings. You will have to zoom in/out on your browser to adjust this. This aspect ratio you can discover on this site. If anyone knows a solution for this I will be grateful.
It’s Brillant Maze Generation! Thank you very much!
I have just ported it to Python mode: http://www.rosettacode.org/wiki/Maze_generation#Processing_Python_mode
I am optimizing the variable names and adding comments for better understanding Also I am taking out some leftover garbage. So maybe you want to check this later on.
OK! Count on me, I’ll have a look.
@jeremydouglass @villares Were you aware of processing ‘plasma effect’ example? I have ported it to JRubyArt Plasma effect - Rosetta Code. I have also ported colored bars task.
Cool! I’ll try my hand at porting 
@noel, @villares I’ve just posted a somewhat object orientated version of the Sierpinski Pentagon on rosetta code. Possibly another task to try?
@paulstgeorge You still want to join us? This is a nice one?
Because this is the message I got from discourse:
This topic is clearly important to you – you’ve posted more than 28% of the replies here.
It could be even better if you gave other people space to share their points of view, too. Can you invite them over?
@Noel The churlish bot has made a mistake. Its logic only works if there is a fixed amount of total space. Only then could you give others space by contributing less.
Anyway, I have looked again and now I know the task I want to tackle: https://www.rosettacode.org/wiki/Combinations.
I assume it is cheating (and boring) to import a library that does the task.
@noel @jeremydouglass @villares The Sierpinksi arrow head curve is missing for processing and python. I’ve just published the JRubyArt version. For the benefit of @villares heres a sierpinski lsystem in processing.py I did earlier.
My second sketch in P5.js. Somewhat concise?
But now I remember; it has to be in P5.java isn’t it?
const s = {x: 20, y: 30, angle: 60};
function setup() {
createCanvas(450, 400);
background(0, 0, 200);
angleMode(DEGREES);
stroke(255);
sc(7, 400, -60);
}
function sc(o, l, angle) {
if (o == 0) {
const {x, y, angle} = s;
s.x += cos(angle) * l;
s.y += sin(angle) * l;
line(x, y, s.x, s.y);
} else {
sc(o-1, l/2, -angle);
s.angle += angle;
sc(o-1, l/2, angle);
s.angle += angle;
sc(o-1, l/2, -angle);
}
}
Did you succeed?..
Your Sierpinski p5.js version w/ less lines: 
"use strict";
const t = { x: 20, y: 30, a: 60 };
function setup() {
createCanvas(450, 400);
background(0, 0, 200).stroke(255).angleMode(DEGREES);
sc(7, 400, -60);
}
function sc(o, l, a, s = t) {
if (o) {
sc(--o, l *= .5, -a).a += a;
sc(o, l, a).a += a;
sc(o, l, -a);
} else line(s.x, s.y, s.x += cos(s.a) * l, s.y += sin(s.a) * l);
return s;
}
I guess no one is going to beat that. 
Maybe just removing"use strict"
It isn’t strict necessary is it?
Googled a bit, and it’s said that it’s good to get rid of many “silent” errors that can occur .
Is that it?