Re: Potential use case for opaque space bank: domain factored network stack

On Mon, 2007-01-08 at 16:45 +0100, Pierre THIERRY wrote:
> Scribit Tom Bachmann dies 08/01/2007 hora 15:31:
> > But, if we, as free software enthusiasts, think that any use of
> > createYield (that is, starting a program you are not allowed to look
> > at) is bad, why should we include that operation into the interface at
> > all?
> 
> It may be because it doesn't really introduce the ability to use hidden
> code and data, because this ability is already present by the way of
> services started on resources we don't have authority to inspect.

Two other reasons:

1. It would allow HURD-NG to directly use the existing Coyotos
framework, which would be a big boost.

2. When the implementation is handled this way, it becomes a very simple
matter to #ifdef-out the createYield operation, while retaining a great
deal of code compatibility.

The discussion at hand does not raise any objection to the constructor
as a means of performing process construction. It is only the opacity
issue that is a concern.
-- 
Jonathan S. Shapiro, Ph.D.
Managing Director
The EROS Group, LLC
+1 443 927 1719 x5100