Inefficient revision tracking of items with 100+ related ited

Summary When ever an item is updated a revision is created containing ONLY first 100 items of a o2m alias relationship , in order to set the data column for that revision. As far as I can tell they aren't being used when viewing past revisions so just take up data space in revisions and increase the time to create a revision. Basic Example We have three primary collections, events and registrations and surveys. Both surveys and registrations are specific to an event Events have alias fields in order to view all the registrations and surveys made for that event. When ever an event is updated a revision is created containing the first 100 registrations and first 100 surveys it finds, in order to set the data column for that revision. This of course is useless then since it doesn't ever contain the full set, but also that revision data would rather be on the opposite of the relationship. It also takes up a lot of space for simply just changing one field on the event.…