Missing getters and setters for flags and inv_dt0 (for serialization purposes) about playrho HOT 1 CLOSED

I wouldn't advise serialization/deserialization while the world was locked. That'd mean copying the world in the middle of a simulation step. Maybe someday that'd be doable, but I think the work for that is significantly harder than dealing with a copy outside of the step. So e_locked state should basically not be an issue.

The e_substepping and e_stepComplete state are a similar situation in that it should basically not be an issue either. Substepping a world is better thought of as a way for debugging what's going on during a world step. And when sub stepping has gotten to the point where an entire step has been sub-stepped through, one gets to the e_stepComplete state. My advice, like with e_locked, pass on this state for now and just don't use sub stepping or don't serialize/deserialize with sub stepping turned on.

I'll have to think about e_needsContactFiltering more.

Hmm... perhaps m_inv_dt0 should not be part of the comparable state of a world used for assessing equality.

In any case, as you're aware, it will be set in the step. And you can set StepConf::dtRatio too. If needed, I believe one could call Step with a deltaTime of zero and the dtRatio of what the previous time delta would have been, multiplied by the m_inv_dt0. See Step(World& world, Time delta, TimestepIters velocityIterations, TimestepIters positionIterations) for example.

from playrho.