[cap-talk] What sustained interest in capabilities

[Mitsu Hadeishi]

On Jan 8, 2009, at 12:00 AM, David-Sarah Hopwood wrote:

> In what sense are any of the most successful computing platforms, such
> as Windows, Unix, and C, themselves adapted to any particular  
> context in
> which they are used?

Again, you're thinking in terms of solving the problem of traditional  
operating systems.  The problem space changes radically when you  
consider the case of building service-oriented "operating systems", so  
to speak.  The layer becomes, effectively, the operating system for  
programs written to the interface of the capability secure layer.  The  
fact that the layer happens to be implemented using legacy code is  
simply an implementation detail.  It would be perfectly possible to  
reimplement the layer on top of a capability secure OS in the future,  
and the only thing that would change is the definition of the legacy  
service capability objects; i.e., for example, a database capability  
wouldn't wrap ACL credentials, it would simply be directly a database  
capability.  Thus, for clients of the system there'd be no change  
whatsoever; it would only be admins who ordinarily would create these  
capabilities who would notice any change.

Mitsu