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

From Tom Lane
Subject Re: roundoff problem in time datatype
Date
Msg-id 24340.1127744911@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
Re: roundoff problem in time datatype
List pgsql-hackers
Dennis Bjorklund <db@zigo.dhs.org> writes:
> On Mon, 26 Sep 2005, Tom Lane wrote:
>> No, I think the usual notation for a leap-second is '23:59:60'.
>> We do allow 60 in the seconds field for this purpose.

> Yes, and it can go up to 23:59:60.999999 (depending on how many fractional 
> seconds one want).

That's an urban legend.  There never have been, and never will be, two
leap seconds instituted in the same minute.  We really should reject
anything larger than '23:59:60'.

> One "solution" is to round '23:59:59.9'::time(0) up to '00:00:00'.

7.2 did that, and we concluded it was broken.
        regards, tom lane


pgsql-hackers by date:

Previous
From: Dennis Bjorklund
Date:
Subject: Re: roundoff problem in time datatype
Next
From: Dennis Bjorklund
Date:
Subject: Re: roundoff problem in time datatype