From: Achilleas Mantzios <achill@matrix.gatewaynet.com>
Organization: Dynacom Tankers Mgmt
To: pgsql-admin@postgresql.org
Subject: Re: how to get a number of seconds in some interval ?
Date: Tue, 13 May 2008 09:56:50 +0300
User-Agent: KMail/1.9.7
Cc: Tom Lane <tgl@sss.pgh.pa.us>, Julius Tuskenis <julius@nsoft.lt>
References: <482849B0.6050103@nsoft.lt>
	<200805121801.35101.achill@matrix.gatewaynet.com>
	<20080.1210604951@sss.pgh.pa.us>
In-Reply-To: <20080.1210604951@sss.pgh.pa.us>
MIME-Version: 1.0
Content-Type: text/plain;
  charset="utf-8"
Content-Transfer-Encoding: quoted-printable
Content-Disposition: inline
Message-Id: <200805130956.51872.achill@matrix.gatewaynet.com>

=CE=A3=CF=84=CE=B9=CF=82 Monday 12 May 2008 18:09:11 =CE=BF/=CE=B7 Tom Lane=
 =CE=AD=CE=B3=CF=81=CE=B1=CF=88=CE=B5:
> Achilleas Mantzios <achill@matrix.gatewaynet.com> writes:
> > =CE=A3=CF=84=CE=B9=CF=82 Monday 12 May 2008 17:32:39 =CE=BF/=CE=B7 Juli=
us Tuskenis =CE=AD=CE=B3=CF=81=CE=B1=CF=88=CE=B5:
> >> do you know why division of intervals was not implemented? As I see it=
 -=20
> >> there should be an easy way to do so. '1 hour 30 minutes' / '1 hour' =
=3D=20
> >> 1.5 .... Seems straight forward...
> >>=20
> > No idea why. However as you pose it,it makes sense.
>=20
> No, it doesn't really, because intervals contain multiple components.
> What would you define as the result of
> 	'1 month 1 hour' / '1 day'
> bearing in mind that the number of days per month is not fixed, and
> neither is the number of hours per day?
>=20

SELECT extract(epoch from '1 month'::interval)/(60*60*24) as "Days in 1 mon=
th interval";
 Days in 1 month interval
=2D-------------------------
                       30
(1 row)
So it seems that the arbitary assumption of (1 month=3D30 days, and 1 day =
=3D 24 hours) has already been made=20
in the extract function.

> You can derive a number if you're willing to make arbitrary assumptions
> about the relationships of the different units (as indeed extract(epoch)
> does), but hard-wiring such assumptions into a basic arithmetic operator
> doesn't seem like a good idea to me.
>=20
> 			regards, tom lane
>=20


=2D-=20
Achilleas Mantzios