I've discovered some odd things about the way elevated event attendee data is stored, and I'm a bit confused as to the business case; wondering if it's a purposeful design decision, a flaw, or something else.
Consider a household with two members, on a system where all contributions and ticket orders are owned by the household; initiator used to indicate which affiliate made the contribution. If I create a contribution for this household on an elevated event, the TX_EVENT_EXTRACT row will be tied to the household ID, but the TX_EVENT_GUEST will NOT be tied to the individual. Instead, the guest row will have the ID of the household, but a guest_name for the A1 household affiliate. Adding the A2 affiliate as another guest means the two guest IDs will be one household and one individual.
This seems to be a design decision that enforces every event_extract has an event_guest for the same ID. This seems to run contrary to the household model, and also causes issues when an invitee makes a decline-with-contribution in response to the invite. The event_extract should still exist, but there is really no event guest here -- nevertheless the guest is required, and so we end up making a "not attending" level to select for the guest. Also reduces flexibility since it's impossible to remove a guest from the guest list when their listed ID is the same as the event owner.
Any comments or clarifications?