From: Nem Tudom <ellenallhatatlan@gmail.com>
Message-ID: <5a4de43a-22ef-4b35-9ea4-57a5b2f520c6@gmail.com>
Date: Tue, 28 Jan 2025 14:00:20 +0000
MIME-Version: 1.0
User-Agent: Mozilla Thunderbird
Subject: Re: Issues with EPOCH-s, TIMESTAMP(TZ)-s and leap seconds.
To: Thomas Munro <thomas.munro@gmail.com>
Cc: pgsql-general <pgsql-general@postgresql.org>
References: <c3520c84-33d3-47f2-a859-091f369bdc91@gmail.com>
 <CA+hUKG+awD-hDwp5bL2kLfv_5oVPd6AoWbN4=qEKiRZkzX7ttQ@mail.gmail.com>
Content-Language: en-US
In-Reply-To: <CA+hUKG+awD-hDwp5bL2kLfv_5oVPd6AoWbN4=qEKiRZkzX7ttQ@mail.gmail.com>
Content-Type: text/plain; charset=UTF-8; format=flowed
Content-Transfer-Encoding: 7bit
Archived-At: <https://www.postgresql.org/message-id/5a4de43a-22ef-4b35-9ea4-57a5b2f520c6%40gmail.com>
Precedence: bulk


On 28/01/2025 00:05, Thomas Munro wrote:


Thanks to you and all the others who took the trouble to reply,


> I showed the bones of how you could do this in SQL here:
> https://www.postgresql.org/message-id/CA%2BhUKGLU9Don4YHnfdzn0eeWQsUu8GJDaLiUAefLLT6%3DmmeGoQ%40mail.gmail.com


The technical explanation is much appreciated - I'd gathered (more or 
less) as much from my searching and reading the leap second Wiki.

So, the situation is that, basically, leap seconds are "fudged" to use 
the technical term!

I asked the question with a view to having accurate TIMESTAMP 
differences - i.e. to the second. However, since everyone is fudging 
(incl. AFAICS Oracle and SQL Server), this means that accepting the 
status quo will just make my inaccuracies will be the same as everyone 
else's, ergo I'm golden!

At least that's one issue that I can safely ignore - I didn't fancy 
implementing this on my own.

Thanks again and rgs,


E!