On Wed, 2006-05-17 at 22:41 +0200, Marcus Brinkmann wrote:
> At Tue, 16 May 2006 00:05:24 +0200,
> Pierre THIERRY <[EMAIL PROTECTED]> wrote:

> > Could you also tell me what you think about the link between
> > virtualization and constructor, in the light of my proposal on this
> > issue? Do you still think, if constructor is implemented the way I
> > describe it, that they are incompatible (and I agree with you that
> > virutalization is indeed a very important feature, for many reasons)?
> 
> As far as I understood your proposal, the disclose flag needed to be
> set voluntarily by the creator of the constructor.  That does not
> satisfy my concerns.

Marcus:

The discussion of the disclose flag has no bearing on virtualizability.
The object is virtualizable or it is not. Whether it is technically
feasible to front-end the object has absolutely nothing to do with
whether its implementation is disclosed.

I understand that you want disclosure to be involuntary. However, we
should not become confused between ideological requirements and
functional requirements.

Jonathan



_______________________________________________
L4-hurd mailing list
[email protected]
http://lists.gnu.org/mailman/listinfo/l4-hurd

Reply via email to