[Sis-csi] RE: networking details

[Lloyd Wood]

At Friday 09/02/2007 15:21 -0800, Adrian J. Hooke wrote:
>At 01:37 PM 2/9/2007, Dave Israel wrote:
>>I think the question is still "Why are we debating the link layer again when people are still not convinced that there is value in basic networking in the first place?"

and we're debating via an email mailing list, which networking makes possible. Some irony there, surely?

>Dave: we are discussing the Link layer precisely because of that very point. Namely, a very large number of people across the international community simply aren't buying into the basic need for networking in space.
>
>The whole reason for raising this issue again was to point out that the real world is rarely black and white. Individuals who bloviate about the immaculate properties of black or the divine nature of white are unhelpful. They polarize opinion and they turn off the vast majority of people who see the shades of gray. The real world of space missions is populated mainly by pragmatic and conservative engineers who will accept change when they are convinced of the need for change, not because some self-appointed guru lectures them that it will be good for their souls.
>
>One way that you convince people to migrate from black to white is to offer hybrid solutions that allow them to transition through gray in an evolutionary manner. The Link layer is important because it needs to facilitate such evolution. We have existing CCSDS Link layers that can equally well support black, white or gray traffic. They are layered,

On that point I must take issue. CCSDS isn't layered. Read:

ftp://ftp-eng.cisco.com/lwood/cleo/hogie-papers/README.html
K. Hogie, E. Criscuolo and R. Parise, <http://dx.doi.org/10.1016/j.comnet.2004.08.005>Using standard Internet Protocols and applications in space, <http://www.elsevier.com/locate/comnet>Computer Networks, special issue on Interplanetary Internet, vol. 47 no. 5, pp. 603-650, April 2005. 

> they are robust, they were specifically designed for space mission use, they are well understood, they are extensively used,

extensively? IP technologies are _extensively_ used. What does that make CCSDS? CCSDS isn't as widely understood as other technologies by as many people as say IP.


>they are well supported by industry 

details, please. From where I'm looking, CCSDS is not well or widely supported.


>and they are agnostic to the technologies above them. They can handle bitstreams or octet streams or TDM or Ethernet frames or Space Packets or IPv4 or IPv6 or ISO-8473 or LTP/DTN or whatever 

ISO-13239 is a very important 'whatever'. That, combined with CCSDS, is the hybrid solution of which you speak.

The thing your evolution metaphor forgets is that evolution often has dead ends.

L.



>- all mixed together on a single Physical channel if desired - and they simply don't care.
>
>The case for basic networking is almost certainly going to be made in the international community by offering people an evolutionary path. The existing CCSDS Link layers are the platforms on which the CCSDS Cislunar working group in the CCSDS Space Internetworking Services Area is going to build its case. End of subject. If people want to propose new Link layers, they can make their separate cases to the CCSDS Space Link Services Area Director.
>
>Best regards
>Adrian
>
>Adrian J. Hooke
>Chairman, CCSDS Engineering Steering Group (CESG) 
>_______________________________________________
>Sis-CSI mailing list
>Sis-CSI at mailman.ccsds.org
>http://mailman.ccsds.org/cgi-bin/mailman/listinfo/sis-csi