The Role of Satellite Services in Internet Event Notification
Michael Gorlick, The Aerospace Corp., firstname.lastname@example.org
[not avail. online, http://www.ics.uci.edu/IRUS/wisen/program.html]
He basically had the same observations. He said that while they
can't get the same time resolution they were talking about
for a global clock, that the resolution for GPS was more than
good enough. I wish I had the numbers and example.
> That's interesting, since it may change quite a bit the way we think
> about distributed system. The basic premice upon which all the theory
> of DOS is based is the absence of global clock, hence impossibility
> to do global synchronization. BUT with the ability to locate yourself
> precisely, a decent (predictable timeouts lenght) and a very reliable
> clock, doing soft synchronization (loose synchronization opposed to to hard
> synchronization e.g. #SYNC line of copper on the motherboard goes to 0)
> becomes theorically valid (I remember one of these DOS a couple of years
> ago using GPS for synchronization, but at that time the cost was out
> of reach for most systems).
> Now after nearly 20 years of failure of research in the domain
> of distributed operating systems, time may have come to remove the
> basic rule who made them impossible to build and start getting something
> that will actually work within 5 years.
> For those interested it may be time to start a PhD on DOS, it may be
> more interesting than 5 years ago ...
> Daniel.Veillard@w3.org | W3C MIT/LCS NE43-344 | Today's Bookmarks :
> Tel : +1 617 253 5884 | 545 Technology Square | Linux, WWW, rpm2html,
> Fax : +1 617 258 5999 | Cambridge, MA 02139 USA | badminton, Kaffe,
> http://www.w3.org/People/W3Cpeople.html#Veillard | HTTP-NG and Amaya.