ISOGR control body: support invalidation/supersession independently

[ronaldtse]

Use cases:

1. Invalidate an entry without any supersession or retirement
2. Invalidate and supersession
3. Supersession without invalidation

Ping @maccraymer to see if this is accurate.

[strogonoff]

Mike & others have confirmed this on the previous meeting. The scenario appears to not be very common, but it’s possible. I believe there is some concern about the semantics of invalid items: if they are indeed added in error then they might not be fit for any further changes.

Another aspect that I recall has been discussed at the same time is gradual supersession over time.

IIRC both features are possible in EPSG and that’s why they were suggested. I’m not sure if we are gunning for feature parity with EPSG, though—19135 compliance should probably be of higher priority.

Filed paneron/registry-kit#30. @ronaldtse there’re some questions in that ticket that you perhaps would know off the top of your head, e.g. would it contradict 19135 (in current or next version) if an already invalid item can be superseded, or an already superseded item can be superseded with more items later.