Date: Tue, 3 Aug 2004 14:50:43 +0300 (EEST)
From: Achilleus Mantzios <achill@matrix.gatewaynet.com>
To: Michael Glaesemann <grzm@myrealbox.com>
Cc: "Pedro B." <pedro.borracha@netcabo.pt>,
	<pgsql-sql@postgresql.org>
Subject: Re: Datetime
In-Reply-To: <9F40F448-E540-11D8-8966-000A95C88220@myrealbox.com>
Message-ID: <Pine.LNX.4.44.0408031449340.7249-100000@matrix.gatewaynet.com>
MIME-Version: 1.0
Content-Type: TEXT/PLAIN; charset=US-ASCII

O kyrios Michael Glaesemann egrapse stis Aug 3, 2004 :

> 
> On Aug 3, 2004, at 7:23 PM, Achilleus Mantzios wrote:
> 
> > now() returns the current UNIX (your running UNIX right?) timestamp 
> > which
> > in turn is
> > measured in seconds,miliseconds since the epoch.
> > i.e. 1970-01-01 00:00:00
> 
> I believe this is incorrect. I believe PostgreSQL uses its own 

Do you suggest postgresql has any other means of getting
time except the time(2) syscall??

> timestamp datatype internally (which is, indeed, not as text in an 
> easy-to-read form). On my machine (running cvs-head),
> 
> test=# select now();
>                now
> -------------------------------
>   2004-08-03 20:27:18.822646+09
> (1 row)
> 
> which is definitely not seconds.milliseconds since epoch. You can use 
> extract to get seconds.milliseconds from epoch, but I don't think this 
> is how it's stored internally.
> 
> test=# select extract(epoch from now());
>      date_part
> -----------------
>   1091532506.3222
> (1 row)
> 
> Just some additional trivia, current_timestamp is an SQL-spec compliant 
> alias for now(), and might be a better choice if one is concerned with 
> compatibility.
> 
> Michael Glaesemann
> grzm myrealbox com
> 
> 
> ---------------------------(end of broadcast)---------------------------
> TIP 5: Have you checked our extensive FAQ?
> 
>                http://www.postgresql.org/docs/faqs/FAQ.html
> 

-- 
-Achilleus