Re: Ada Operating System

["Marin David Condic" <dont.bother.mcondic.auntie.spam@[acm.org>]

It would be interesting to build an OS that used some sort of
command/response protocol with message passing and synchronous/asynchronous
operation. Clearly, with the underlying mechanism invisible to the
application, it wouldn't know if it made a request of the OS on the local
chip, another chip on the bus or a processor clear across the country.

The downside of it is that it would be inefficient for the cases where the
requests were local. Some analysis and modeling would need to be done to
determine if, given modern microprocessor speed, the penalty would be too
extreme to warrant its use.

It might be possible to provide *two* API's for the OS. One with standard,
synchronous  subprogram calls and another that duplicated the functionality
with command/response messaging. But utilizing it (from a compiler
perspective) might be way too complicated.

MDC
--
Marin David Condic
Senior Software Engineer
Pace Micro Technology Americas    www.pacemicro.com
Enabling the digital revolution
e-Mail:    marin.condic@pacemicro.com

"James Ross" <rem.jr@rem.webross.com> wrote in message
news:gr10auodfr11kgtci3pvtj9dufc17r9tn5@4ax.com...
> On Mon, 25 Mar 2002 17:25:53 GMT, Darren New <dnew@san.rr.com> wrote:
>
> >Imagine an OS where all IPC is done via
> >message passing of messages between queues.
>
> This is an interesting concept. If "all" OS calls used such a
> mechanism, then any application written is a target for a client /
> server scenario without any redesign (or recompilation for that
> matter).  The OS could simply forward the calls to the server running
> the app.  It wouldn't matter if the box responding to the call was
> your own or across the country!
> JR
>