The matching functions is where the EDPS can be a real improvement over Reflex / OCA-rules. Or it can limit things, because the matching function is defined on the (calibration) product that the system is looking for, while in OCA-rules they are defined on the (processed) product that needs input. Therefore it seems that it has become more difficult to use different selection criteria for different final products. E.g. sometimes we might want to focus on astrometry at the cost of photometry, and sometimes the other way around; this seems harder in the EDPS than with OCA-rules. We should experiment more.
Hugo Buddelmeijer authored