home *** CD-ROM | disk | FTP | other *** search
- Path: sparky!uunet!cs.utexas.edu!zaphod.mps.ohio-state.edu!saimiri.primate.wisc.edu!ames!agate!ucbvax!DIKU.DK!thorinn
- From: thorinn@DIKU.DK
- Newsgroups: comp.protocols.time.ntp
- Subject: XNTP3/MIPS RISCos 5.01 solved (ish)
- Message-ID: <9301251139.AA08448@tyr.diku.dk>
- Date: 25 Jan 93 11:39:40 GMT
- References: <9301251022.aa12468@salmon.maths.tcd.ie>
- Sender: usenet@ucbvax.BERKELEY.EDU
- Distribution: inet
- Organization: The Internet
- Lines: 20
-
- Alan,
-
- from your log it looks like the system time jumped one second backward
- in the minute after 05:56:26. Perhaps the SETTIMEOFDAY_DAMAGED patch
- does not always guess correctly to which side RISC/os 5 is going to
- jump?
-
- The compliance is not really the compliance any more --- it is log2 of
- the loop time constant in seconds. The actual value, as a time
- interval, may have some significance in phase locked loop theory, but
- in the NTP loop filter it is just a number that is used to scale
- certain other numbers --- the larger the time constant, the
- ``stiffer'' the loop, and the longer the poll interval.
-
- The compliance used to be another quantity with a dimension of time, a
- long-term decaying average of the loop offset, scaled to ``the same''
- units as the loop time constant which it was used to derive. (Smaller
- compliance => larger loop time constant.)
-
- Lars Mathiesen (U of Copenhagen CS Dep) <thorinn@diku.dk> (Humour NOT marked)
-