Pre-tabulated environments generated directly from C8 Media
The WIP Cherenkov MR !333 cannot use the C8 media/atmosphere for photon propagation - it has to use one of several dozen pre-tabulated atmospheres (from C7). This creates a disconnect where the atmosphere that the shower "sees" is different than the atmosphere that the photons "see" - currently, the burden is on the user to understand and try and replicate the C7 atmospheres in C8. This also means that Cherenkov can only be run in a specific set of atmospheres/geometries à la C7.
This is far from ideal and (I think) will severely limit C8 down the road. I think one of the biggest advantages in C8 over C7 is the ability to create custom atmospheres and geometries as there are many experiments that don't fit in the "standard" down-ward air shower picture (ANITA, POEMMA, PUEO, EUSO, TRINITY, BEACON, and the list goes on)
In the radio MR, we directly use the C8 media interfaces and don't use pre-tabulated environments for this reason (even though it is currently much slower). Similarly, now "radio photons" will see a different atmosphere than "optical photons" which is ridiculous.
The best solution (I think) would be to tabulate the C8 atmosphere before the start of a library and cache it (with some unique hash or ID based upon the atmosphere parameters). This may be expensive to do once but could then be loaded from the cache for future simulations. We can then create an C8 media interface that is backed by this tabulated atmosphere and can be used by radio, Cherenkov, fluorescence, etc.
I'm concerned about merging the Cherenkov MR since it naturally creates that "fork" in the media system. Does anyone have any objections to implementing a proposal like this?