Hello all!
I have an application in which I have to track all group-membership history of multiple groups, each group comprised of multiple members. Think of it as fraternity membership in a college: many fraternities, many members in each.
Group-members are either added or deleted, never updated. A given member however can leave and then rejoin later its former group or another group. Typically, more then one member is joining and leaving at a given date.
And a key requirement is to be able to reconstruct the group membership as it was at any past date. (The time granularity is days.) So far it seems that I am talking about a classical "Point-in-Time Architecture" or PTA.
After searching these forums and the web, I came across two approaches:
These two approaches deal only with single rows, at least in their examples. The question is whether or not either one can deal with retrieving a multiple members group, given the Group-ID and a (past) date.
Any ideas?
Regards,
Meir