[ntp:questions] TSC, default precision, FreeBSD
Richard B. Gilbert
rgilbert88 at comcast.net
Sat Sep 5 19:15:55 UTC 2009
Garrett Wollman wrote:
> In article <9fa237cd-eaf8-4303-8720-3e290547fa90 at g1g2000pra.googlegroups.com>,
> Dave Hart <davehart at gmail.com> wrote:
>> I hope this help you understand NTP precision. I also hope you will
>> cease attempting to conflate it with any other definitions of
>> "precision" or related terms.
> It's a pity that the authors of the NTP specification chose to
> Humpty-Dumpty an existing word with a well-defined meaning.
At least it's stated what meaning it was "Humpty-Dumpty'd" to.
Precision is the smallest possible difference between two consecutive
readings of the system clock. It's normally expressed as a negative
power of two. I once wrote a tiny program to read the clock twice in
succession and do the arithmetic and ran it on a Sun Ultra 10
workstation with a 440 MHz CPU. IIRC the result was 2^-21.
It has nothing whatever to do with accuracy, it simply specifies how
finely you can slice the time.
More information about the questions