[Sis-csi] noting unusual CCSDS security critique

[Krupiarz, Christopher]

Lloyd,

First, I apologize for forwarding your email to the mailing list.  As your response was a clarification as to where I could find the CFDP reference, it seemed relevant to the rest of the group and I thought it was innocuous enough to forward, but you are correct that was not proper etiquette.

As for CFDP itself, I'd say we simply reached a different conclusion on its complexity (although I appreciated the chance to learn a new UK colloquialism in "fiddly" :).  We've had success with it on MESSENGER, appreciated that it had gone through a rigorous international standards practice, and would certainly consider it again for future missions.  However, at the end of the day, we all have to make choices which enable the best data return for our missions.

As this is probably more relevant to the CFDP mailing list, I'd suggest we continue this there if you are interested.  Feel free to forward my comments to that list.  I'd also encourage you to post your concerns regarding CFDP there as well.  We all learn in exchanging our successes and, in particular, our failures.

Chris



At 03:05 PM 10/5/2005 -0400, Krupiarz, Christopher wrote:
>Lloyd,
>
>Thanks for the clarification.  I'm not sure if this is a question for
>the SSTL authors instead of you, but I'll ask anyway.
>
>In the most recent paper, the replacement protocol (Saratoga) is
>described as having a smaller footprint and increased performance
>compared to the CFDP implementation and that the CFDP implementation was
>"considered large, slow and resource-hungry".

Christopher -- as you have read and as you state, that paper clearly refers 
specifically to the CFDP _implementation_. Twice.

It would be unscientific to draw conclusions about the worth of the CFDP 
protocol from a single implementation alone. However, reading the relevant 
Blue and Green Books should be more than sufficient to convince anyone that 
CFDP is large, fiddly, and has an overly complex state machine.

Saratoga ('a custom rate-based UDP-based file transfer protocol designed 
and implemented by SSTL', as the paper says) is proprietary to SSTL, as is 
the reference given for that. There are many other proprietary UDP-based 
transfer protocols (Real, for example).

In future, please do not forward off-list mail to a mailing list.

L.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mailman.ccsds.org/pipermail/sis-csi/attachments/20051005/61d1646d/attachment.html