[BUG]

[cdaniluk]

I don't think that PR #4 actually solves the problem. I noticed in the project that you used this in, you computed route_tables_count based on the sum of subnets, not the sum of route tables. I suspect that the route table count itself is truly non-deterministic at the time of plan.

[smiller171]

@cdaniluk I think #4 is the only way to solve this problem, by allowing you to manage the count logic outside this module in a way that is deterministic for the specific use case in the parent module.

IMO if I messed up the logic here a bug should be opened in that repo. I don't think it's possible to solve this in a cleaner way in this repo