I'd like to be able to select a world, select chunks to be rendered, and initialize render settings, and render an image either via command-line options or via some simple scripting language and without every opening the UI.
I started attempting to make this work but all the rendering methods are so closely tied to the UI that I didn't manage to get it to render an image and write it to disk. I figured I should at least ask you how feasible this would be before going and making a lot of changes that you might not want to merge back to the master branch.
(In case you're wondering, the use case for this, and also the reason I added the parallel projection, is to use Chunky to generate tiles for an overhead map.)
I've been thinking about using a regular text file to store scene properties, this way all the setup could be done by a simple script, then sent to Chunky on the command-line. I think the most natural thing to do is use a regular .properties file, a standard that should work well.
I added an issue for loading scene settings from a properties file here: #51
Since version 1.2 the scene description format is JSON based. I think that this should make it sufficiently simple to set up scenes via scripting.