Doing it better 1999-02-08 g
###

From Reuben.

> This affects the decision of how low to put the VM, relative to the
> communication primitives. Not a lot, but it does. Bear it in mind.

Not really. You can rewrite either the VM (which doesn't really exist in
my system, which is just a collection of objects) or the communications
primitives in either virtual or native code. The main thing my system
does, I suppose, is provide that interface.

> Another interesting idea. Doesn't it mean that you are throwing away
> useful information, though?

No, you can add any information you need later.

> > No, I'm avoiding it.
>
> Then you are underestimating it :-)

No, I'm postponing it. Like with Mite. I think I've got a bit of a feel
for postponing-without-implication.

> Ever heard of natural selection? Rabbits? I want a bit more power. Maybe
> I'm paranoid.

You can do that via the object/process model: it's perfectly possible to
build objects/processes so that you can tell them to die and they will.

> Absolutely. A time-slice resource could be an abstract datatype, just like
> memory.

As to memory it's interesting to see that some people (e.g. the Sombrero
OS) think that you should have a single address space even across a LAN,
and then you can do without IPC and messaging, and just use subroutine
calls.

>  - Implementing it on top of an accepted VM.
>  - Implementing an accepted VM on top of it.

!

--
http://www.cl.cam.ac.uk/users/rrt1001/ | maxim, n.  wisdom for fools