[coyotos-dev] Thoughts on (non)persistence
Jonathan S. Shapiro
shap at eros-os.org
Thu Jan 27 07:42:06 EST 2005
On Mon, 2005-01-24 at 14:12 -0800, Norman Hardy wrote:
> On Jan 24, 2005, at 10:55 AM, Jonathan S. Shapiro wrote:
> > On Mon, 2005-01-24 at 07:10 -0800, Charles Landau wrote:
> >> Jonathan wrote "I've put a bunch of background thinking into what is
> >> lost when transparent machine-wide persistence is removed from the
> >> architecture". It would add clarity to this discussion if he would
> >> say what he's replacing it with.
> > I completely agree. If I knew the answer, I would post it. This is the
> > question we are now trying to resolve. My sense at the moment is that
> > we
> > are moving toward something akin to the E VAT persistence model of
> > sturdyrefs and liverefs, and no consistency guarantees beyond
> > individual
> > objects.
> Individual objects? What is an Individual object?
This isn't productive. What needs to happen is for us to get the Coyotos
spec captured so that we all know what we are reacting to and
discussing. Answering questions piecemeal just increases the load on
As I said in mail earlier today, we do not plan to do orthogonal
persistence in Coyotos.
More information about the coyotos-dev