Is there a means of differentiating rows added after a certain date without explicitly storing that information as a column item? I working with a database that is constantly being filled with information and I'm trying to devise a means of efficiently working with a row subset defined as newly added rows, but I need that date meta information to make said differentiation possible. Is this at all possible?
A dirty solution, that sadly only works on some tables, I've thought up includes keeping track of the number of rows I've collected (this works because the primary key is generated by the database) and then making queries with an offset. If the date added meta information exists then I can devise a generic solution that will suite all of my tables.
Please note that I can't add an explicit data added column because I already have a ton of rows collected over several years.