Mikey Nichols

Thanks for commenting! I’m happy that it resonated =) I think having that short-tick vs long-tick test is going to save me later too.

That’s a really good test case. It hits the exact boundary resolve_hit() is supposed to unify, instead of testing clicks and ticks separately. A click leaving the slime at 1 HP, then a one-second tick finishing it, would be a nice guard against missed transitions or double rewards as this grows. I’m probably adding that one.

Yeah, I think that’s the cleaner way to handle it. Keep tick() deterministic, apply the away-time cap once per session, then pass the credited seconds into the simulation. That also keeps the cap and the simulation easy to test separately. Good catch.

Exactly! That separation ended up being one of my favorite parts of this step. I started from the much smaller problem of “how do I test an hour away without waiting an hour?” and game.tick(3600) fell out of that. But it also creates that nice boundary you described: whatever is running the game can measure wall time, while the Rust side only has to care about what that amount of elapsed time means for the game state. I like the comparison to separating simulation from render frames too. That is pretty much where this series is heading next — pulling the game rules away from the environment that happens to be running them.