To: gdavis@refractions.net
cc: pgsql-sql@postgresql.org, pgsql-hackers@postgresql.org
Subject: timestamp subtraction (was Re: formatting intervals with to_char)
In-reply-to: <4522EC42.1080500@refractions.net> 
References: <4522EC42.1080500@refractions.net>
Comments: In-reply-to Graham Davis <gdavis@refractions.net>
	message dated "Tue, 03 Oct 2006 16:03:30 -0700"
Date: Thu, 05 Oct 2006 12:50:54 -0400
Message-ID: <21619.1160067054@sss.pgh.pa.us>
From: Tom Lane <tgl@sss.pgh.pa.us>

Graham Davis <gdavis@refractions.net> writes:
> The documentation for to_char states that:

> "|to_char(interval)| formats HH and HH12 as hours in a single day, while 
> HH24 can output hours exceeding a single day, e.g. >24."

> However I can not get it to work with time intervals that span more than 
> 1 day.

Well, it does in fact print intervals exceeding 24 hours:

regression=# select to_char('48 hours'::interval, 'HH24:MI:SS');
 to_char
----------
 48:00:00
(1 row)

However, '48 hours' and '2 days' are not the same thing.  The problem
with the case you give is really that timestamp_mi applies justify_hours
to its result --- that is, 

regression=# select ('2006-09-15 23:59:00'::timestamp - '2006-09-01 09:30:41'::timestamp);
     ?column?
------------------
 14 days 14:28:19
(1 row)

should be reporting '350:28:19' instead.

This is a hack that was done to minimize the changes in the regression
test expected outputs when we changed type interval from months/seconds
to months/days/seconds.  But I wonder whether it wasn't a dumb idea.
It is certainly inconsistent, as noted in the code comments.

I'm tempted to propose that we remove the justify_hours call, and tell
anyone who really wants the old results to apply justify_hours() to the
subtraction result for themselves.  Not sure what the fallout would be,
though.

			regards, tom lane