discussion:Where do we go from here?

[Gary D. Duzan]

In Message <Pine.SUN.3.91.941107151701.11360D-100000@crl.crl.com> ,
   Mike Prince <mprince@crl.com> wrote:

=>
=>
=>On Mon, 7 Nov 1994, Raul Deluth Miller wrote:
=>
=>> Raul Miller:
=>> . > also: machine abstraction suitable for describing system
=>> . > capabilities [for estimation and planning purposes]
=>> 
=>> Obviously, we're not going to even have good estimates all the time
=>> (on what kind of resources we're going to need), but we should be able
=>> to act smart with decent estimates.
=>
=>I'm very interested in discussing this further.  Perhaps each object 
=>could provide info about it's functionality.  Device specific objects 
=>would be "tied" to a particular machine and could not migrate.  Thus 
=>you'd make a grocery list of needs, and look for a workspace having all 
=>the objects that couldn't be imported in.  The rest could be brought in 
=>and away you'd go.
=>
   I've thought about this some myself. Some of this might be handled
by language-level hooks, which can get messy if you try to do it
automatically, but programmer hints like "this is a CPU-sucking loop"
or "major I/O here" would be a start. Beyond that, for programs that
are run a number of times, collecting statistics on execution behavior
may help predict what it will do in the future.

                                      Gary D. Duzan
                         Humble Practitioner of the Computer Arts