Re: Translucent storage: design, pros, and cons

On Fri, 2007-01-12 at 20:13 +0100, Tom Bachmann wrote:
> -----BEGIN PGP SIGNED MESSAGE-----
> Hash: SHA1
> 
> Jonathan S. Shapiro schrieb:
> > On Fri, 2007-01-12 at 19:58 +0100, Tom Bachmann wrote:
> > 
> >> Well. I have the binary of a program (or, actually, it's source code,
> >> but there exists a compiled version, for convenience). I load it into
> >> memory and give it whatever capabilities I think are needed, enforcing
> >> whatever properties I want. What can the constructor do that I can't?
> > 
> > Tom: 
> > 
> >   1) What about capabilities it holds that you do not?
> 
> I create it. All capabilities that it has have been given to it by me.
> This is, as I understand it, a core property of the partial design
> Marcus described in part two of his notes.

Why do you assume that you create it? You may supply the storage for the
instance but not the constructor.

> 
> >   2) There is still utility in doing this by means of a commonly used
> >      routine.
> 
> Yes, and as I have said before, I don't argue that it is to be removed
> per se (actually, not at all, for exactly the reason you state), but
> that it is not to be considered a "relevant" part of the design, if it's
> utility drops to only being commonly used for convenience (which is what
> it appears to me to be in the translucent storage / hierarchic trust
> relationships design).
> - --
> - -ness-
> -----BEGIN PGP SIGNATURE-----
> Version: GnuPG v1.4.6 (GNU/Linux)
> 
> iD8DBQFFp93nvD/ijq9JWhsRAof6AJoD62rlDqVLKWETwmVh1Jtzkgr9GgCfW8MU
> rbpH4tD0M/JnFPDf12mqpN8=
> =ZJrj
> -----END PGP SIGNATURE-----
> 
> 
> _______________________________________________
> L4-hurd mailing list
> [email protected]
> http://lists.gnu.org/mailman/listinfo/l4-hurd
-- 
Jonathan S. Shapiro, Ph.D.
Managing Director
The EROS Group, LLC
+1 443 927 1719 x5100