Right now if a science frame doesn't have any calibration data with it when the entire program ID dataset is downloaded that science observation will be skipped. This is essentially because the calibrations directory wouldn't be created. However, we might be able to make this more robust by adding a failover option here. It remains to be seen how many programs this affects, or even if the right choice here is to use calibrations from a day ahead or behind.
Right now if a science frame doesn't have any calibration data with it when the entire program ID dataset is downloaded that science observation will be skipped. This is essentially because the calibrations directory wouldn't be created. However, we might be able to make this more robust by adding a failover option here. It remains to be seen how many programs this affects, or even if the right choice here is to use calibrations from a day ahead or behind.