IPC and SRP...

Johan Fabry johan.fabry at vub.ac.be
Mon Aug 28 16:16:32 UTC 2000


btanksley at hifn.com wrote:

> >I've been following this discussion, since I'm working on something
> >similar myself. My aim is also to get different VM's to
> >communicate on a
> >somewhat higher level than having to code with sockets, however instead
> >of IPC, I am working with different computers. But since I've noticed
> >that you concluded you would have to use sockets, my system could be of
> >some use to you.
> 
> >BTW, CORBA 1 and 2 have no serialisation, except for basic types (ints,
> >strings, ...) and datastructures you can define in IDL (structs, enums,
> >...). If I remember correctly, serialisation *is* planned for CORBA 3
> 
> It's finished, well-defined, and implemented in a number of places.  If I
> were you I would work hard to support it; there's no reason I can see to
> reinvent the wheel.

Remote Message Passing is a lightweight protocol by design, by
comparison CORBA is heavyweight. Although RMP is inspired by CORBA, it
is not CORBA compliant. I dont completely reinvent the wheel though,
just the parts that make RMP easier on the user ;-)

This is because (IMO) CORBA is not really easy to learn/use and quite
complicated. RMP was designed to be easier to learn/use than CORBA and
to be simple. This has as an obvious result that RMP is Squeak-only, but
I feel that it is a constraint I can live with for now ;-).

(Also I do not have the time/resources to make RMP CORBA compliant.)

--
"You are more than the sum       Johan Fabry - Johan.Fabry at vub.ac.be
 of what you consume.            Vrije Universiteit Brussel
 Desire is not an occupation."   Programming Technology Lab, Room 10F709
  -- KMFDM                       Pleinlaan 2, 1050 Brussels, Belgium





More information about the Squeak-dev mailing list