It's common for tables to store start and end dates to record a time period. But this brings a challenge:

How do you ensure the end of one period is the start of the next? i.e. there are no gaps between periods and no periods overlap?

This session contrasted having start and end dates columns versus a single event date column and generating the periods.

Highlights include:

00:00:40 - Using foreign keys to ensure consecutive start-end dates
00:03:00 - Single change date column and generate start-end period
00:05:00 - Definition of consecutive periods
00:09:20 - Discussion & vote for one column vs start/end dates
00:13:25 - Data changes with one column vs start/end dates
00:19:30 - Demo of data changes with one column vs start/end dates
00:31:30 - Null vs magic dates for the end date for the current row
00:43:10 - Temporal validity to simplify start-end date queries
00:43:50 - Demo of null vs magic dates for the end date for the current row
00:50:30 - Performance tuning date queries
00:51:00 - Index column order for start date, end date columns
00:56:00 - Performance of queries to find the active row with one column
00:58:00 - Demo of query performance
01:07:00 - Session recap