if you do find that the overhead is too large, please file an issue and we can see about either optimizing that further or adding a batch check