Thread: Precision of time types

Hello,

and yet another clueless newbie question:

The precision of time types is specified as follows in the
documentation:

"For the time types, the allowed range of p is from 0 to 6 when
eight-byte integer storage is used, or from 0 to 10 when floating-point
storage is used."

When/how is the internal storage method determined?

Compile-time option?
Database-initialisation time option?
Other?

Unfortunately, the schema I am trying to implement using PostgreSQL
specifies 9 digits (nanosecond precision) for its custom datetime
column (don't ask me why, I'M just a poor mech eng).

So, do I need to use a numeric type for the seconds and define my own
composite datetime column or can I use a properly parameterized
built-in time type of PostgreSQL?

TIA,

Sincerely,

Wolfgang Keller

On Nov 13, 2007, at 6:45 , Wolfgang Keller wrote:

> When/how is the internal storage method determined?
>
> Compile-time option?

Yes, depending on the --enable-integer-datetimes flag.

> So, do I need to use a numeric type for the seconds and define my
> own composite datetime column or can I use a properly parameterized
> built-in time type of PostgreSQL?


Are you looking at timestamps or durations? If you're looking at
timestamps, you could use timestamp(0) + an additional column to hold
the number of nanoseconds. If durations (mis-named intervals in SQL),
you could either store it as an interval(0) + an additional column
for nanoseconds, or perhaps the total number of seconds in a numeric
column.

As you suggest, you could wrap the two columns into a single
composite type.

Michael Glaesemann
grzm seespotcode net