Not much more to be said, really, apart from the fact that as I updated the code from a hacky-mess to something not quite as hacky with error checks etc, speed did drop sufficiently that I think soft-edge shadows probably isn’t the way to go. Hard-edges though are still nicely fast 
How’d you do that?
The tile layer data lets you know if there’s a tile there or not, so you can find edges that face away from the light source. Shadow edges are just a line from the wall’s corner away from the light. So to answer your question, mostly it is just maths
Very cool.
rakoonic,
Thanks for all the demos and inspiration over the years. Any chance some of your experiments might end up as a plugin in the Corona Marketplace?
-Jonathan
Possibly.
At the moment this is just a proof of concept test, it would need a fair amount of work to be useable which requires time. Time which I hope to have in the next few weeks or so
How’d you do that?
The tile layer data lets you know if there’s a tile there or not, so you can find edges that face away from the light source. Shadow edges are just a line from the wall’s corner away from the light. So to answer your question, mostly it is just maths
Very cool.
rakoonic,
Thanks for all the demos and inspiration over the years. Any chance some of your experiments might end up as a plugin in the Corona Marketplace?
-Jonathan
Possibly.
At the moment this is just a proof of concept test, it would need a fair amount of work to be useable which requires time. Time which I hope to have in the next few weeks or so