branchrust_reboot/main/industrial_perf_improvementscancel
Today
Some server industrial conveyor performance improvements -Cache at the start of a run if an output has empty slots, that way if an item isn't stackable we can immediately rule out the entire container instead of iterating over the whole thing -Compare slots by item id rather than object comparer -maxItemStacksMovedPerTickIndustrial convar is now enforced across an entire conveyor move operation rather than per input (this will slow industrial systems down, we may want to evaluate raising the default convar from 12) These changes take a problematic conveyor configuration from 670ms to 12ms processing time