Is there any documentation how fiber positions and other aspects of focal plane geometry (e.g. exact location and geometry of the AG cameras) will be communicated between software components? If so, I'd be grateful for a pointer!

I am afraid it is captured in libraries. I have - for a long time - on my table to dig through those.

Sorry, not very helpful. I think Moritani san can give the initial pointers.

Based on some recent additional requests, it seems that I do not only need the AG camera footprint on the focal plane, but also information about the number and arrangement of pixels for each camera. I'd like to avoid having to access this kind of information within shuffle, since it requires interfacing with more components and I do not understand the added value, but as things currently are, I will need to obtain this from somewhere.

The AG camera part is done, which should be the most important one. Cobra locations are needed only for diagnostic plots and not for shuffle's central task, which is to identify guide star candidates.

Do we still need the Cobra locations for this piece of code? Currently we don't output any plots, and if this is OK with all users, this issue can be closed.