mayer at gis.net
Tue Jun 14 06:51:40 PDT 2005
----- Original Message Follows -----
> > If you don't believe
> > that is an accurate characterization of system performance, you
> > should take the issue up with the real physicists at NIST, USNO, PTB
> and NPL.
> No problem - its not. Its not the physicists that need to characterize
> the performance of NTP, its the OS platform and its loading and
> fingerprints that are the issue.
If you want to know things like accuracy, you DO need to talk to the
physicists. It's not the only thing you need to worry about but it
is one of them. O/S platform characterization is also important since
the ability of each platform to process packets, clocks, handle
interrupts, other processes running on the O/S and hardware and for
that matter the hardware configuration all matter. Whether you
use IDE or SCSI disks, for example can make a big difference,
> What I am saying is that NTP becuase its used in
> Evidentiary Testimony of legally binding events is in a different
> league when it was just you and some other physicists experimenting
> with how to synchornize distributed computing networks...
Yes but are you going to do this for people who can guarantee that
noone is going to be able to touch their systems without their knowing
and that nothing but what is specified in the analysis is what runs
on the machine in question?
More information about the hackers