[3622] in Release_7.7_team
Re: Outreach to multi-partition Sun users
daemon@ATHENA.MIT.EDU (Jonathon Weiss)
Mon Dec 9 02:52:37 2002
Message-Id: <200212090752.CAA25076@bearing-an-hourglass.mit.edu>
From: Jonathon Weiss <jweiss@MIT.EDU>
To: "t. belton" <tbelton@MIT.EDU>
cc: Bill Cattey <wdc@MIT.EDU>, release-team@MIT.EDU
In-reply-to: Your message of "Fri, 06 Dec 2002 12:51:36 EST."
<Pine.GSO.4.33L.0212061245500.14676-100000@iphigenia.mit.edu>
Date: Mon, 09 Dec 2002 02:52:34 -0500
> So, yes, without a certain amount of lead time to move that stuff, it
> would indeed "inconvenience" me if Athena were suddenly reinstalled as a
> single partition.
Just to clarify, we're talking about whether it is acceptable to
require a re-install of multi-partition machines before they can take
9.2. We are not talking about doing that automatically, or suddenly.
> curious about this announcement in a philosophy-of-life kind of way. If we
> are trying to move more and more things local, aren't we removing the
> whole point of having developed a massive, successful, distributed
> computing system in the first place?
One of the general ideas behind distributed computing is to distribute
the work of computing to the end machines. By puting copies of
various pieces of regularly used software onto client workstations,
we're actually making that better, since the workstation won't have to
talk to a central server to get the program (while the user is
waiting). We are, in fact, enhancing the distributed computing
system, in order to remove it we owuld have to do things like put user
spcific data on workstations, and require that users use specific
machines to do specific things (thus making each workstation a central
server of a sort). If it helps, think of the local copy of the
software as a cache of what is in AFS, and remember that AFS is the
location of record for this software.
Does that help?
Jonathon
