In your simulation, are you doing the computation work (ie computing how agents react) and the…

Yes indeed. Thus the interest in the nearly orthogonal approach of OofA .. i.e. an agent set as an object with a typed array per variable .. or possibly a typed array as a set of variables like x,y,z. A “position” variable.

I’ve even build an OofA module once, with some nice sugar like “push” with automatic extension of the TA’s sizes. This offered the user to either directly access the arrays if wanted, but also provided an “object illusion” if needed.

I did this initially so that we could have huge number of agents .. TAs reduce the size considerably. Then webgl popped up and naturally TAs are a win there. Workers too sound interesting but I haven’t tried that yet. I should tho.

And given wasm on the horizon, typed arrays are even more interesting.

Hmm..maybe I oughta dust that OofA module off for yet another nifty story!

Like what you read? Give Owen Densmore a round of applause.

From a quick cheer to a standing ovation, clap to show how much you enjoyed this story.