Time variance

From Wikipedia, the free encyclopedia

[edit] Time Variance

Time variance is the ability to remember historic perspectives. The requirement is to be able to know how something was classified or who owned something and how this changed as time passed.

[edit] Understanding Time Variance

Future change, be it organizational, regulatory or geographical is hard to conceive. In 1988, who imagined that within a few years, Yugoslavia and East Germany would cease to exist? Enabling a data warehouse to report pre- and post-change information together in a meaningful context is very expensive and time-consuming. Couple that with the pressure to rapidly meet other business requirements, and with the inability for any of us to predict change (especially at system design time!), and you can see why the time-variant reporting requirement is often ignored. But at what expense? The real-life case below illustrates the value of time-variant reporting: A beverage company paid rebates to customers at year-end, based on ownership of customer sites at year end. The sales data warehouse did not reflect customers selling sites to one another throughout the year, resulting in mis-payments and a multi-million dollar customer service dilemma. For regulatory compliance and other reasons, data warehouses must remember how things were in the past because at some point business people will expect to be able to be report on them that way.