Re: Ada.Calendar and NTP (and Unix Epoch)

[Simon Wright <simon@pushface.org>]

erlo.haugen@gmail.com writes:

>> My gut feeling is that if NTP gives you a time in the range 1-1-1900
>> to 12-31-1900, something is pretty wrong because computers
>> didn't exist during that time period.  So I'm not really
>> sure what the issue is.  I think you need to be more specific about
>> what you're trying to do.  What kind of value is your program
>> using for an "NTP time"--is it an integer, and if so, what
>> does it represent (seconds since 1-1-1900, nanoseconds since
>> 1-1-1900, or what)?  Are you trying to see if an NTP time is less
>> than or greater than a value of Ada.Calendar.Time, or compute the
>> difference in seconds (or something else) between an NTP time value
>> and an Ada.Calendar.Time value, or what?
>> 
>>                         -- Adam
> NTP epoch 1 started 1-1-1900 0:00, just like the Unix epoch started
> 1-1-1970 0:00. So when I get a time from an NTP-server, I get seconds
> from the start of the NTP epoch.

Yes; but if you regard the 64-bit value as a *signed* value then the
current NTP value is about -24 years. The next NTP epoch is in 2036,
which is easily representable in Ada.Calendar.Time. So, *add* -24 to
2036 and, presto, 2012! In UTC.

Of course, if you are worried about leap seconds, you are in for a world
of pain either way.

From Wikipedia[1],

   NTP delivers UTC time. UTC is subject to scheduled leap seconds to
   synchronize the timescale to the rotation of the earth. When a leap
   second is added, NTP is suspended for 1 second. Because NTP has no
   mechanism for remembering the history of leap seconds, leap seconds
   cause the entire NTP timescale to shift by 1 second.

[1] http://en.wikipedia.org/wiki/Network_Time_Protocol#Leap_seconds