My first positive structural change to NTP

Eric S. Raymond esr at
Sun Jun 26 16:29:42 UTC 2016

Achim Gratz <Stromeko at>:
> Eric S. Raymond writes:
> > The reason I disagree is I think you're overfocusing on the fact that
> > both refclocks are the same physical device and underfocusing on the
> > fact that they're two different data channels, possibly with different
> > fudges and modes.
> No, it's exactly my contention that this is an implementation detail
> that ntpd doesn't really need to care about.

Sorry, I think you have that *exactly* backwards.  From the point of
view of the time-sync algorithms everything is sequences of timestamps
pairs with statistics attached to each sequence.  The fact that
multiple timestamp-pair streams may be supplied by a single device is
the implementation detail.  I think the config language should reflect
this view.

If you want to persuade me that the new syntax should be something other
than the extremely simple change I have described, you'll have to do
more than talk theory.  Write up an actual proposal as Daniel Franke
has done for the restrict replacement, describing syntax and semantics.
Explain why it should be preferred.

Be aware that I will consider excessive complexity a crash landing and
keep it simple. I'm not interested in replacing the obscure with the
more obscure.
		<a href="">Eric S. Raymond</a>

More information about the devel mailing list