"When?" It's a basic question that we ask all the ... time.

Stenn, Harlan

When did "it" happen? When will we do "it"? At face value, this is pretty simple and easy stuff. For many social interaction, getting the time right to within a few minutes is just fine. There are many billing applications that need better accuracy, as they charge in six-second

increments. How important might accurate time be in some medical and health care situations? In the final 9 seconds before the total power failure that hit the US Northeast and Canadian East coast in 2003, about 18,000 events happened in the power distribution grid. The clocks used

to record these events were pretty much unsynchronized, and nowhere near usefully accurate. So the logged information was pretty much useless.

The "When" question is answered with a "timestamp". A timestamp might be a note in your calendar (Dinner at 6pm), or the "Date" entry in an email message.

We've all been using a very simple and basic timestamp to record the time since long before we had computers. We have the knowledge and experience to engineer, implement, and deploy a much better timestamp. High-quality timetamps improve the usefulness of data. Better data means things like advances in health care, the ability to make better

decisions, better efficiency, reduced liability, and better

profitability and financial management.

This talk will explore some of the problems with existing timestamps and the consequences of these problems. You'll learn about NTF's solution to these problems, its General Timestamp API and Library, and how it can help.

Return to invited speaker list