:
:On Tue, 13 Jul 1999 11:59:25 -0700 (PDT) 
: Matthew Dillon <dillon@apollo.backplane.com> wrote:
:
: >     We could have the ability to mark processes as being more or less
: >     preferable as kill candidates.  I'm not sure I really care anymore,
: >     though... there is so much disk space available now that it is fairly
: >     difficult to run the system out of swap space.  I don't think I've 
: >     run any of my personal systems out of swap space for at least a year 
: >     now!  Usually the biggest process is the one responsible (note: MFS
: >     processes do not count, and they are immune from being killed).
:
:...I suppose it depends on what market you're going for, too.  Some
:systems (not even necessarily OLD systems, but maybe modern, embedded
:ones, too) don't always have the option of having "so much disk space
:available".
:
:Seems like you want your operating system to behave the `correct' way
:depending on the environment in which it's being used.
:
:        -- Jason R. Thorpe <thorpej@nas.nasa.gov>

    If you don't have the disk necessary for a standard overcommit model to
    work, you definitely do not have the disk necessary for a non-overcommit 
    model to work.

    If you were to try to implement overcommit protection on a system
    with insufficient disk space, it probably wouldn't have enough reservable
    space to run basic system daemons, much less any actual servers.

    Read the part again where I mentioned the swap requirements for a 
    non-overcommit model to operate at the same level as the current model.

					-Matt
					Matthew Dillon 
					<dillon@backplane.com>