Results for perpetually observable and un-observable positions

[jwfraustro]

Reading through the text, I have been unable to find a description for how a service should handle positions that it is not capable of observing, or will always be capable of observing. Both cases apply to ground-based observatories especially.

For positions an observatory will never be able to observe, it makes sense that the query simply returns no results. Should there be any other indication?

For positions that are always observable, what are the values of t_start , t_stop and max_obs a service should return? It makes sense to simply return the current time of the query for t_start.

This has been addressed as part of the MR, and is no longer an issue:
Perhaps this ties into the question of default TIME parameter behavior? Would it be sensible for a service to simply return the maximum TIME in the future they are willing to provide? For services that choose not to implement a maximum TIME limit, what should the value be?