XPost: sci.space.history   
   From: pfeiffer@cs.nmsu.edu   
      
   Rick Jones  writes:   
   >   
   > When you are talking about addressing at the IP level, do you mean   
   > things like IPv4 and IPv6 addresses?  I wouldn't think that there   
   > would be many of those assigned to a spacecraft at all - perhaps one   
   > or two.   
   >   
   > Using "pure" IP comms though would bring one up against the 8-bit   
   > protocol field of IPv4 (never can remember if that was increased for   
   > v6) for routing things within the spacecraft.  256 "addresses" within   
   > the spacecraft seems a little future-limiting.  I would think you   
   > would want a ULP above IP - say UDP or something else to give broader   
   > mux/demux opportunities.  Does CCSDS have even fewer bits for demux   
   > than the 8 in IP(v4)?   
      
   I guess I don't see why you'd only have a small number of nodes on the   
   spacecraft, nor why you'd use the protocol field for routing within   
   the spacecraft.  Rather, I'd expect every logical node in the   
   spacecraft to have an IP address, and reasonable protocols (like UDP,   
   for instance) to talk to all of them.  Quite possibly with address   
   translation so there's only externally visible address.   
      
   Of course, all this presupposes that IP makes sense in deep space   
   anyway.  I know nothing about deep space protocols, but I sure haven't   
   seen anything in this thread to make IP look like a good idea!   
   --   
   Joseph J. Pfeiffer, Jr., Ph.D.       Phone -- (505) 646-1605   
   Department of Computer Science       FAX   -- (505) 646-1002   
   New Mexico State University          http://www.cs.nmsu.edu/~pfeiffer   
      
   --- SoupGate-Win32 v1.05   
    * Origin: you cannot sedate... all the things you hate (1:229/2)