[CESG] Magenta Book issue: "Profiles"
Chris.Taylor at esa.int
Chris.Taylor at esa.int
Mon Feb 7 07:17:21 EST 2011
I have two points to add to the discussion:
What ever is finally decided we shouldn't end-up with a proliferation of
CCSDS books. I am really against this on two accounts: the effort required in
an environment of ever decreasing budgets, and the added complexity of having
to refer to multiple books and explain to users how everything fits together.
Defining individual protocols as separate books provides flexibility but it
does make the process of assembling a profile for a mission much more
difficult. At the moment we have books that define services, protocols and
stacks (AOS blue book), Services and one of more protocols (SPP, TM),
services only (SOIS subnetwork). We also have Green books which explain how
things fit together. I am rather in favour of the status quo, maybe with some
minor alterations. I see nothing wrong, and very much support, defining
classes of protocol in a single document as in the present CFDP book.
What I would strongly object to is any retroactive change to the status and
process of the published and under preparation SOIS books. We have had many
discussions on whether these should be Blue or Magenta and in the end we
homed in on Magenta.
Regards,
//ct
From: "Hooke, Adrian J (9000)" <adrian.j.hooke at jpl.nasa.gov>
To: "Shames, Peter M (313B)" <peter.m.shames at jpl.nasa.gov>, CCSDS Engineering Steering Group - CESG Exec
<cesg at mailman.ccsds.org>
Date: 04/02/2011 23:26
Subject: RE: [CESG] Magenta Book issue: "Profiles"
Sent by: cesg-bounces at mailman.ccsds.org
Peter: I think that I agree with you. To mark the event, maybe we should take
the rest of the week off?
I do think that the finer-grained “Adaptation” and “Utilization” categories
are useful. I never liked the term “Application Profile” because it’s too
easy to confuse how the protocol is being applied with what the protocol does
: that’s because the word “Application” has another meaning (as in
Application Layer protocol).
I also agree strongly with your assertion that if a specification is
implementable, does have interoperability properties, and clearly needs to be
tested, then it’s a Blue Book. Mike Kearney said this another way in a
private note to me earlier, in the context of interoperability testing:
I would advocate that regardless, we get to the point where Blue Books
*always* require prototyping and Magenta books *never* require
prototyping. Having a state where Magenta books sometimes do and
sometimes don’t require prototyping would be a bad thing. And I still
like the idea of reading the book, and if you think it requires
prototyping, then it must be blue… if not, it must be magenta.
So the litmus test between Blue and Magenta appears to be as stated above: it
is implementable, it does have interoperability properties, and it clearly
needs to be tested.
If that’s true, then within the Blue category do we have the following? –
CCSDS Recommended Standard (something that internally contains a native
specification developed by CCSDS)
CCSDS Recommended Standard: Adaptation Profile (something that adopts/adapts
a native specification contained somewhere else)
CCSDS Recommended Standard: Utilization Profile (something that specifies how
to use one or more Blue Books to perform a particular function)
It’s interesting that we seem to be massaging our “adopt, adapt, develop”
mantra here, by squeezing it out to “adopt, adapt, develop, use”. The word
Adaptation seems particularly right, since one definition is: a written work
(as a novel) that has been recast in a new form; "the play is an adaptation
of a short novel".
///a
From: Shames, Peter M (313B)
Sent: Friday, February 04, 2011 11:30 AM
To: Hooke, Adrian J (9000); CCSDS Engineering Steering Group - CESG Exec
Subject: Re: [CESG] Magenta Book issue: "Profiles"
Dear Colleagues,
In the new "new document “ORGANIZATION AND PROCESSES FOR THE CONSULTATIVE
COMMITTEE FOR SPACE DATA SYSTEMS”", the Draft CCSDS A02x1y2a (attached),
there is considerably more text about “Application Profiles”, including
discussion of the need for prototyping these "stacks of protocols. The
reasoning behind this is that these types of documents do, in fact, describe
something that is intended to be interoperable, for example. "IP Over CCSDS".
Since this is the case, we then need to diligently carry out the same sort of
point by point interoperability tests that we do for our usual, single
layered standards.
While editing this revision of the document, long overdue to be published, it
became necessary to add the clause, in Annex B, pgs B-6, Sec B2.3a:
Application Profiles, in particular, may require prototype
implementations and interoperability validation. Whereas these practices
do not, in and of themselves, define any new standards, they do define
how a set of standards are to be “stacked” in order to provide
end-to-end communications services, and this functionality and correct
operation must be validated. In these cases a sufficiently detailed test
plan must be developed, as would be done for a Blue Book, and
interoperability testing must be performed and documented in a Yellow
Book test report.
Some of you may recall that we originally created the notion of Application
Profiles and Magenta Books precisely to deal with the fact that they are not
the same as our usual, normative, One-ISO-Layer-at-a-Time, Blue Books. Over
time we have identified other types of Magenta Books, and now it seems that
we are trying to distinguish other types of Blue Books, those that are
"Adaptations" of other standards, either developed in CCSDS or external by
external standards bodies, and those that are "Utilizations" of other
standards. These distinctions appear to be useful ones, and I agree that
MTS, LTP, and DVB-S fit into something we could call an Adaptation category
and that IP Over CCSDS fits into what we could call a Utilization category.
In fact, one of our foundational Blue Books, CCSDS 401.0-B-20, "Radio
Frequency and Modulation Systems—Part 1: Earth Stations and Spacecraft", has
for many years been exactly a Utilization style document. Much of this
document consists of references to external standards, particularly for
modulation, (NRZ, BPSK, GMSK, OQPSK, etc), and it does not define any of this
within the body of the document, just how, when, and where these external
standards are to be utilized.
Where I differ from Adrian, if we are going to introduce this new terminology
and the associated distinctions into the "revised yet one more time"
ORGANIZATION AND PROCESSES FOR THE CONSULTATIVE COMMITTEE FOR SPACE DATA
SYSTEMS, is in whether we should take the position that only "Adaptations"
are Blue. Pretty clearly Adaptations will still need to have
interoperability testing done, will need two independent implementations,
will need to have two separate exercises of any required licensing, and will
also need to have all of their clauses tested.
The exact same situation appears to be the case for Application Profiles, and
other "Utilizations". So, rather than having Application Profiles, or
"Utilizations" in general, continue to be a special case of Magenta Books,
the only one that requires interoperability testing, I would recommend that
we declare that all Applications Profiles, and "Utilizations" in general, are
Blue Books. They are implementable, do have interoperability properties, and
clearly need to be tested. This would tend, I believe, to simplify the
whole CCSDS A02x1y2 document structure and make the distinctions between Blue
and Magenta rather clearer, since there would no longer be a need to explain
why this class of Magenta Books, and only this, need implementations and
interoperability testing, but none of the others do.
Bottom line: I support Adrian's proposal but would take it one step further,
to add both Adaptation and Utilization categories to the Blue Book track.
Regards, Peter
From: Adrian Hooke <Adrian.J.Hooke at jpl.nasa.gov>
Date: Fri, 4 Feb 2011 07:24:55 -0800
To: CCSDS Engineering Steering Group - CESG Exec <cesg at mailman.ccsds.org>
Subject: [CESG] Magenta Book issue: "Profiles"
Area Directors: as you will have observed in recent days, we are having some
problems deciding whether documents should be categorized as Blue or Magenta.
The latest problems have surfaced with several documents that point to
external specifications in a normative way (e.g., SOIS-MTS, SIS-LTP,
SLS-DVBS) rather than containing their own normative specifications. This is
sometimes done for reasons of copyright, i.e., we can’t simply cut-and-paste
the text of another standard and make it our own. However, in all of these
cases one thing is clear: we are pointing to another normative specification
and adopting all of it, or a precise subset/superset of it, as our own
normative technical standard. So is it a Blue Book or a Magenta Book?
Some have argued that we are essentially creating an “Application Profile”
and as such the document should be Magenta. However, in the current governing
documents http://public.ccsds.org/publications/YellowBooks.aspx (the
“Procedures Manual for the Consultative Committee for Space Data Systems” and
the “Restructured Organization and Processes for the Consultative Committee
for Space Data Systems”) there is no such formal document category as an
“Application Profile Magenta Book”. The Procedures Manual refers – once – to
“application profiles for CCSDS specifications recommended for use in
particular mission support configurations”. The restructuring book notes –
once – that “a Recommended Practice might specify some specific “Application
Profiles” of multiple CCSDS Standards that are recommended for use in
particular mission support configurations”. In the still-draft new document
“ORGANIZATION AND PROCESSES FOR THE CONSULTATIVE COMMITTEE FOR SPACE DATA
SYSTEMS” there is a bit more proposed text about “Application Profiles”, but
we haven’t agreed on that yet.
Clearly, we need some better guidance to help distinguish between a
Recommended Standard and a Recommended Practice in the cases where we are
adopting a normative “profile” of another standard. So it may be helpful to
invent language to distinguish between:
a) a profile that specifies how to accomplish a new general
function by normative reference to specific normative text in some
other specification(s) [e.g., LTP, MTS or DVB-S] and;
b) a profile that specifies how to support a particular mission
class by recommended reference to how some other specification(s) may
be used or stacked together [e.g., IP-over-CCSDS].
We could call the former an “Adaptation Profile” that would live in a Blue
Book and the latter a “Utilization Profile” that would live in a Magenta
Book.
As an interesting aside, it looks like we in fact already have got
internationally standardized “Adaptation Profiles” that have been certified
Blue: the four Service Classes in section 7 of the CFDP Blue Book are exactly
and precisely Adaptation Profiles. We just happen to have packed them into
727.0-B-4 along with everything else related to file transfer, instead of
publishing each one in its own (thin) Blue Book. And it appears that we’re
doing the very same thing with Conformance Classes in the AMS book.
What’s important here is that by formally recognizing these Adaptation
Profiles as legitimate independent Recommended Standards we would make it
possible to do flexible things (like incrementally adding more Conformance
Classes to AMS or more Service Classes to CFDP) by publishing thin, easily
reviewed new Blue Books as the demands of flight mission operations evolve,
without ever having to go back and hack/reissue the original protocol specs.
There are other issues to resolve with Magenta Books (concerning
interoperability testing), but for now could you please comment on the above
proposal?
///adrian
Adrian J. Hooke
Chairman, CCSDS Engineering Steering Group (CESG)
Space Communications and Navigation Office (SCaN)
Space Operations Mission Directorate
NASA Headquarters
Washington DC 20024-3210
_______________________________________________
CESG mailing list
CESG at mailman.ccsds.org
http://mailman.ccsds.org/mailman/listinfo/cesg
More information about the CESG
mailing list