home *** CD-ROM | disk | FTP | other *** search
- Newsgroups: alt.sources.d
- Path: sparky!uunet!zaphod.mps.ohio-state.edu!cs.utexas.edu!uwm.edu!psuvax1!atlantis.psu.edu!barr
- From: barr@pop.psu.edu (David Barr)
- Subject: Re: adjclock - automatically maintain real-time clock
- Message-ID: <g0g1Hdew?a@atlantis.psu.edu>
- Sender: news@atlantis.psu.edu (Usenet)
- Organization: Penn State Population Research Institute
- References: <9211181646.AA14462@wendy-fate.UU.NET> <x?a1Hcw!za@atlantis.psu.edu> <21904@rpp386.lonestar.org>
- Date: Thu, 19 Nov 92 14:01:27 GMT
- Lines: 16
-
- In article <21904@rpp386.lonestar.org> jfh@rpp386.cactus.org (John F. Haugh II) writes:
- >Is adjclock good enough for Kerberos? Probably. I see plenty of
- >machines running AFS more than 1000ms apart. Is adjclock cheaper than
- >an Internet connection? Definitely.
-
- Who says you need an Internet connection for NTP? I said that clocks
- need to be accurate _relative to each other_, not necessarily absolutely
- accurate to some Stratum 1 time source. Thanks to Sun's horrible clocks
- and terrible console drivers I would see drifts of 2 or more seconds
- in a single day.
-
- --Dave
- --
- System Administrator, Population Research Institute barr@pop.psu.edu
- But you don't cure bad art by locking it all up in one museum,
- you cure it by throwing tomatoes at bad artists. - Barry Shein
-