[ntp:hackers] Leap second support in versions prior to 4.2.6

Greg Dowd gdowd at symmetricom.com
Mon Jun 4 17:48:29 UTC 2012


I think we agree but I'm not sure :-)  In any case, I was not debating the method of leap second handling in NTP.  The method is CAPABLE of being deterministic which is mostly my interest.  However, the implementation chooses not to use that capability which I find unfortunate.  

Greg Dowd
gdowd at symmetricom dot com (antispam format)
Symmetricom, Inc.
www.symmetricom.com
"A clever person solves a problem.  A wise person avoids it." Albert Einstein


-----Original Message-----
From: Mike S [mailto:mikes at flatsurface.com] 
Sent: Monday, June 04, 2012 10:41 AM
To: Greg Dowd
Cc: hackers at lists.ntp.org
Subject: Re: [ntp:hackers] Leap second support in versions prior to 4.2.6

On 6/4/2012 1:15 PM, Greg Dowd wrote:
> The challenge of course, is that the ITU definition only defines a
> leap in calendar format representations of time which is becoming an
> anachronism in some ways.  The method of propagation of leap seconds
> in binary UTC based systems (e.g., NTP) is implementation dependent.
> So, the net result in a binary system is that either you massively
> change clock frequency or you repeat a value.

Which are problems with the system/implementation. I don't see any point 
in debating whether one kludge is better than another on a system which 
can't properly handle leap seconds. Any kludge is bad.



More information about the hackers mailing list