demo: placeholder image generator

[automata]

trafficstars

First study for a placeholder image generator using bezier curves only. Some samples:

[bergie]

@automata looks very nice! We need to be able to store stuff like this as high-res images

[automata]

Thanks @bergie! I guess we can export the canvas as a blob on the runtime.

Now we can use a string as a seed to random numbers, creating the same image for each string (using noflo-seedrandom).

Some examples using bezier curves emerging from figure center:

Those ones are more interesting as "paintings", covering the whole canvas without a reference point:

Which approach is more interesting for placeholder images @bergie, @forresto?

[forresto]

The full-screen ones would fit better, I think.

We should experiment with inputting a primary color, and doing transforms on that, or an array of colors.

Maybe a transform-color module would be good, but make-color outputs strings, which makes that more annoying. Otherwise could just take hue + 180 etc.

I guess css string -> HSL(A) object component would be helpful.

[automata]

:+1:

Another points to consider:

• theme based (e.g. following proper color palettes and shapes)
• use existing images (e.g. nature) to synthesize a more abstract shape using Delaunay/Voronoi tessellation

[automata]

More studies (in higher dimension, using node-canvas).

[automata]

Using noflo-geometry: