Re: Predictable Internet Time

Joe Touch <> Tue, 28 March 2017 19:19 UTC

Return-Path: <>
Received: from localhost (localhost []) by (Postfix) with ESMTP id 80EC612955E for <>; Tue, 28 Mar 2017 12:19:20 -0700 (PDT)
X-Virus-Scanned: amavisd-new at
X-Spam-Flag: NO
X-Spam-Score: -6.901
X-Spam-Status: No, score=-6.901 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, RCVD_IN_DNSWL_HI=-5, RP_MATCHES_RCVD=-0.001] autolearn=ham autolearn_force=no
Received: from ([]) by localhost ( []) (amavisd-new, port 10024) with ESMTP id clOa-YFfARkS for <>; Tue, 28 Mar 2017 12:19:19 -0700 (PDT)
Received: from ( []) (using TLSv1.2 with cipher ECDHE-RSA-AES128-GCM-SHA256 (128/128 bits)) (No client certificate requested) by (Postfix) with ESMTPS id F3D7D129486 for <>; Tue, 28 Mar 2017 12:19:18 -0700 (PDT)
Received: from [] ( []) (authenticated bits=0) by (8.13.8/8.13.8) with ESMTP id v2SJIbD6014358 (version=TLSv1/SSLv3 cipher=DHE-RSA-AES128-SHA bits=128 verify=NOT); Tue, 28 Mar 2017 12:18:38 -0700 (PDT)
Subject: Re: Predictable Internet Time
To: Michael Breuer <>, Patrik Fältstr öm <>
References: <> <> <> <> <> <> <> <> <> <>
Cc: IETF Discussion Mailing List <>
From: Joe Touch <>
Message-ID: <>
Date: Tue, 28 Mar 2017 12:18:37 -0700
User-Agent: Mozilla/5.0 (Windows NT 10.0; WOW64; rv:45.0) Gecko/20100101 Thunderbird/45.8.0
MIME-Version: 1.0
In-Reply-To: <>
Content-Type: text/plain; charset="windows-1252"
Content-Transfer-Encoding: 7bit
X-ISI-4-43-8-MailScanner: Found to be clean
Archived-At: <>
X-Mailman-Version: 2.1.22
Precedence: list
List-Id: IETF-Discussion <>
List-Unsubscribe: <>, <>
List-Archive: <>
List-Post: <>
List-Help: <>
List-Subscribe: <>, <>
X-List-Received-Date: Tue, 28 Mar 2017 19:19:21 -0000

On 3/28/2017 12:09 PM, Michael Breuer wrote:
> ...
>> When UTC adds a leap second, nothing different happens to POSIX
>> time.
> That's a quite complex topic to find precise wording. If the last
> statement would be right, wouldn't this be a contradiction to your
> earlier statement, POSIX time be defined *not counting* leap seconds?

POSIX ignores leap seconds - whether added (typical) or removed (if that
ever happens).

> AFAIK POSIX time is continuous and not counting leap seconds. So when
> UTC adds a leap second, something quite different happens to POSIX time.
POSIX doesn't know or care - it just keeps ticking.

> It will change the leap-second-offset between UTC and Unix time.

That happens because of something UTC did, not something POSIX did.

> To my opinion, any time system utilising leap seconds isn't continuos at
> all. There are continuous time systems like TAI or GPS system time (or
> any other satellite navigation system time) 
GPS reports TAI+19s, but at least one other satellite system jumps with
leaps (GLONASS).

> and POSIX time, which (by
> design) shift away from UTC over time.
TAI does too.

TAI and POSIX also drift away from each other in unpredictable ways
because TAI uses SI seconds and POSIX has no precise definition (it
appears to assume a second derived from a mean solar day, but it isn't
precisly spec'd).