[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [moca] Locations?



* Peter Sewell (Peter.Sewell@xxxxxxxxxxxx) wrote:
> ok then :-) There are (at least) four different notions, none of which
> necessarily coincide, that have been used in various combinations. One
> can have:
> 
> - units of failure (machines or runtime system instances for process
> failure, and/or units within which communication is guaranteed
> reliable if the endpoints do not fail)
> 
> - units of migration (or, more generally, units of running computation
> that can asynchronously be marshalled up)
> 
> - units of trust (both large administrative domains, and small
> critical regions from which a particular secret should not escape. As
> Sanjiva observes, in interesting cases the latter cut across other
> structure.)
> 
> - units of synchronisation (global in the case of pi and djoin,
> partitioned in many other calculi)
> 
> together with some finer variations, eg units of namespace control 
> (the body of which can control the names of sub-locations, as in seal
> and box-pi).

There is also unit of control of communication (as in the M), which is 
slightly related to trust and namespace control.

Alan Schmitt

-- 
The hacker: someone who figured things out and made something cool happen.
  
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
The "models for mobility" mailing list     mailto:moca@xxxxxxxxxxxxxxx
 http://www-sop.inria.fr/mimosa/personnel/Davide.Sangiorgi/moca.html