Re: roundoff problem in time datatype - Mailing list pgsql-hackers

From Tom Lane
Subject Re: roundoff problem in time datatype
Date
Msg-id 29688.1127749780@sss.pgh.pa.us
Whole thread Raw
In response to Re: roundoff problem in time datatype  (Dennis Bjorklund <db@zigo.dhs.org>)
Responses Re: roundoff problem in time datatype
Re: roundoff problem in time datatype
List pgsql-hackers
Dennis Bjorklund <db@zigo.dhs.org> writes:
> Do the sql standard say anything on the matter?

It doesn't seem very helpful.  AFAICS, we should interpret storing
'23:59:59.99' into a TIME(0) field as a cast from TIME(2) to TIME(0),
and the spec defines that as
       15) If TD is the datetime data type TIME WITHOUT TIME ZONE, then let           TSP be the <time precision> of
TD.
           b) If SD is TIME WITHOUT TIME ZONE, then TV is SV, with             implementation-defined rounding or
truncationif necessary.
 

So it's "implementation-defined" what we do.
        regards, tom lane


pgsql-hackers by date:

Previous
From: Tom Lane
Date:
Subject: Re: roundoff problem in time datatype
Next
From: Gaetano Mendola
Date:
Subject: Re: roundoff problem in time datatype