Should I re-compile my games every month?

Should I re-compile my games every two months?

Would the various improvements in optimization that probably occur every so often be worth it? Would it depend if I use physics or not?

Probably. Keep in mind, if you have a really old project -- say, prior to r155 with the introduction of collision cells -- you may actually get a slower result if your collision events aren't top level (ie, if you are trying to narrow down affected objects thru other means).

collision cells -- you may actually get a slower result if your collision events aren't top level (ie, if you are trying to narrow down affected objects thru other means).

Has anyone shown an example .capx/events of this to illustrate it, or has Ashley explained that aspect of it somewhere?

— https://www.scirra.com/blog/ashley/6/collision-cell-optimisation-in-r155

Thanks. It still feels vague though:

[quote:g7ng7pwp]If Is overlapping (or On collision) comes after another condition which has picked certain instances, it can no longer use collision cells.

Does it have to ALWAYS be at the top of an event, or are there exceptions? Does the "picked certain instances" refer only to conditions belonging to the instances in the overlap, or is it conditions picking anything at all (even if the instances weren't included in the overlap)?

IMHO.. IF it works, don't mess with it is a good idea..

So, if your game is made with certain version of Construct (and plugins) keep an archive of those in case you'll need to fix something on later times.