[ntp:questions] Experimental ntpd for Windows

Martin Burnicki martin.burnicki at meinberg.de
Wed Feb 4 09:30:29 UTC 2009


Harlan Stenn wrote:
> I vote we try some things and see what happens and have some productive
> dialogues.

I do partly agree with Dave, and partly with Danny.

We may have some intermediate versions of ntp-dev which can be used to test
a new feature or algorithm on a couple of machines, maybe even with some
restrictions concerning other applications.

However, release (-stable) versions should work as expected regardless which
other apps are running.

Unfortunately you don't know in advance which other apps set the MM timer
high, and AFAIK there's even no general way to check whether any other app
has already set the MM timer high (maybe except for Vista, where the clock
resolution obviously changes from about 15 ms to 1~3 ms if the MM timer is
set high).

IMHO any patch which enhances the clock discipline under Vista is welcome as
long as it doesn't break the clock disciplie for non-Vista versions of
Windows.

Martin
-- 
Martin Burnicki

Meinberg Funkuhren
Bad Pyrmont
Germany




More information about the questions mailing list