I don't know the internals of either. In a Paint Worklet, the canvas is generated for you, and your painter is automatically called every time an input propert

I don't know the internals of either. In a Paint Worklet, the canvas is generated for you, and your painter is automatically called every time an input property changes. In -webkit-canvas, you manually instantiate the canvas and manually repaint it. It's the difference between reactive and procedural programming (even though the actual painter is procedural in either case). Here's a gist I threw together showing how to use the same paint function with either API: https://gist.github.com/appsforartists/e5d2a4b7826bf5962fad142bcd255465 https://gist.github.com/appsforartists/e5d2a4b7826bf5962fad1...