How to handle "non-trivial" scans

At present in the system a lot of the design and examples of configuration and scan setup are centred around an assumption of a simple "scan" on a fixed position on the sky with a scan_id attached to identify that scan and its data.

However, there are/will be other types of "scan" required for observing. Two in particular will be taking short scans offset from a central pointing position in a cross shape ("5-point" being a common example) used for reference pointing observations and raster/on-the-fly scans, for holography observing and later for science. The key question this ADR aims to answer is are these types of "scan" to be considered one scan (with one scan_id) or multiple scans (with multiple scan_ids) and what is the impact of either choice on all subsystems.

Why now? Reference pointing and holography observations will be required at AA0.5.

Scope: All subsystems.

Quality: Architectural integrity