Обсуждение: timestamps-accuracy

We noticed that several records which have a  time stamp column  have
the same time stamp which i can understand given the time stamps have 1
microsecond resolution.

generail this don't hurt us but we have some needs with data coming from
manufacturing testing applications that needs to keep the timestamps to
.000,000,001 aka 1 nano second.
what would be the easiest way to do this?

Just as mind numbing we actual measure electrical resistance out that
far to and have to keep track of it in the database.

On Fri, May 09, 2008 at 04:59:30AM -0400, Justin wrote:
> generail this don't hurt us but we have some needs with data coming from
> manufacturing testing applications that needs to keep the timestamps to
> .000,000,001 aka 1 nano second.
> what would be the easiest way to do this?

I would suggest storing it as epoch time (seconds since X) in a numeric
fields and create some helper functions to get data in and out in the
right format. Or a timestamp(0) column with an additional nano-second
column.

Depending on how often you use it it may be interesting to create a
specific type, but this may come close enough.

Have a nice day,
--
Martijn van Oosterhout   <kleptog@svana.org>   http://svana.org/kleptog/
> Please line up in a tree and maintain the heap invariant while
> boarding. Thank you for flying nlogn airlines.