Excerpts from Tom Lane's message of mar mar 15 11:42:06 -0300 2011:
> "Kevin Grittner" <Kevin.Grittner@wicourts.gov> writes:
> > Tom Lane <tgl@sss.pgh.pa.us> wrote:
> >> Dimitri Fontaine <dimitri@2ndQuadrant.fr> writes:
> >>> Would it help moving toward Leap Second support, and is this
> >>> something we want to have?
>
> >> IMO we don't want to have that, as it would completely bollix
> >> datetime calculations of all kinds. You couldn't even count on
> >> stored timestamps not changing their meaning.
>
> > I'm inclined to agree, but if that's the choice, should we stop
> > claiming that we're using UTC, and instead claim UT1 support? It
> > always seemed a little odd to me that the docs say UTC but there's
> > no actual support for leap seconds in calculations.
>
> Maybe, but if the docs started talking about that, we'd have to define
> the term every time. The number of PG users who know what UT1 is can
> probably be counted without running out of toes.
A small note somewhere visible would suffice: "these docs talk about UTC
but they really mean UT1 because we have no leap seconds support".
--
Álvaro Herrera <alvherre@commandprompt.com>
The PostgreSQL Company - Command Prompt, Inc.
PostgreSQL Replication, Consulting, Custom Development, 24x7 support