On Sat, 19 Nov 2005 19:49:53 -0600, J Moore wrote:
> Nov 19 17:00:34 opie ntpd[6238]: clock error = +26.46s; adjusting... > Nov 19 17:04:03 opie ntpd[6238]: clock error = +17.27s; adjusting... > Nov 19 17:07:31 opie ntpd[6238]: clock error = +9.24s; adjusting... > Nov 19 17:11:52 opie ntpd[6238]: clock error = -0.63s; adjusting... > Nov 19 17:16:10 opie ntpd[6238]: clock error = -4.75s; adjusting... I have no say in it; but I may have an opinion: Yes; go for it. It is comprehensible from the start; and I rather see an estimate of the error; so eventually I can still rdate when off by many hours. It also shows the regulation pretty well, with a slight overshot as to be expected. Just this morning one of my boxes showed Nov 20 09:14:28 mybox ntpd[21036]: adjusting local clock by -459.160367s Nov 20 09:15:31 mybox ntpd[21036]: adjusting local clock by -459.023878s Nov 20 09:16:05 mybox ntpd[21036]: adjusting local clock by -457.982831s Nov 20 09:20:20 mybox ntpd[21036]: adjusting local clock by -457.527840s and I was asking myself: "What does this mean to me ?" after the earlier long exchange stating that is does *not* mean "I adjust by 459... seconds now". So what is it, practically: drift ? calculated compensation value ? clock error is much cleaner, IMHO. Uwe
