The HxRG readout software uses the coordinate system which Teledyne uses, which is as seen from inside the detector looking out. Pretty much everyone is finding this confusing, and wants to change it before any more work is done.

Yes, any software above raw FITS readers can and will flip and rotate just fine. And the H4s are physically rotated such that dispersion is along rows not columns, so for reductions we will still need to adjust the coordinate system.

[ And yes, perhaps the CCD coordinate system should have been flipped so that fiber numbers increase from left-to-right. ]

But since no one is consuming the H4 data yet, there is little harm in flipping X at the lowest level, and the optical engineers will be less likely to make mistakes. Increasing wavelength with increasing X.