This is more of a rant about nuke's paint, or maybe I'm doing something wrong with the paint node, so advice is welcome.
I know that the paint node in Nuke is still considered to be in Beta, but there seems to be a huge feature missing, I would even call it a bug.
When I'm cloning, and I sample an area I cloned in that node nuke doesn't "source" from the painted area (what I see on the screen) it actually sources back from the input image. In other words I can't clone painted areas that have been cloned, very poor paint node implementation. I know this can be worked around by putting another paint node after and then cloning again, but thats just gonna add plenty of paint nodes that could be avoided.
This really sucks because our studio is in the process of transitioning from Shake to Nuke but the current job requires a lot of painting since we didn't get clean plates. We basically had 2 options: Paint in Shake composite in Nuke, or ditch Nuke and stay in Shake. Obviously we decided to stick with Shake for this project.
Don't get me wrong, I like Nuke and I'm happy with it, however there are still some (simple) things where good 'ol 10 year old Shake 2.5 is more effective than Nuke.