[Css-csts] RIDs re BDD Procedure in CSTS FW

[Fred Brosi]

CSTS colleagues,
Several inter-related RIDs regarding the Buffered Data Delivery
procedure (Sec. 4.5)  of the CSTS Framework were assigned to me at the
Fall 2010 WG meeting.  The RIDs concerned the need to distinguish the
data units that are the result of service production (for all instances
of service) from the data units sent to individual users by instances of
a procedure derived from the BDD procedure. In working on those issues,
I discovered other problems in Section 4.5. Some of these are technical
errors, some are ambiguities, some are typographical or formatting
errors. My revised version of the BDD section is show in the PDF file:
FW-BDD-fmb revision-2011.03.24.pdf
The latest version of this section, from Yves' December version, is
shown in:
FW-BDD-921x1r2[Draft 201012.pdf
The comparison of these versions, showing insertions and deletions, is
in:
FW-BDD-Compare-2011.03.24.pdf
These three files, along with the complete FW document Word file with my
revisions:
FW-fmb 921r2[Draft 2011.03.17-rev.doc
Are on the CWE in http://tinyurl.com/48gpdqx
MI think that the motivation for most of my changes will be easy to
understand by comparing the text in the two versions, using the
Word-generated comparison as a guide to where things have been changed,
but I'll provide a few additional comments and examples of problems I
found. 'FW' is used here to refer to the version of Section 4.5 with
which I started.
1.	Service 'production' is sometimes used where 'provision' is
meant, and vice-versa.
2.	The section is intended to specify an abstract procedure from
which a 'concrete', implementable procedure can be derived, but the FW
frequently doesn't make the distinction. 
3.	The FW often uses 'the service' or 'the service provider' when
'the (BDD) procedure' or 'a BDD-derived procedure' or 'an instance of a
BDD-derived procedure' is what is meant. Sometime the text seems to
imply that there is one 'user', but an instance of service might involve
a number of instances of the BDD procedure, some operating in real-time
mode and others in complete mode, each on behalf of different users.
4.	I found a number of statements that are correct only for either
real-time or complete mode, but are written as if they are true for
both.
5.	There are some aspects of the specification that seem to require
understanding of the way SLE-RAF was implemented. This makes the BDD
specification hard to understand, at best, and possibly incomplete or
wrong.
I've tried to resolve these issues, partly through revision of the text,
and also by providing a series of progressively more detailed diagrams.
I've made a few decisions on style and organization that I believe will
make understanding the section easier for a new reader. I think the
revised version is an improvement, but I'm sure it's far from perfect. 
Best regards,
Fred Brosi    +1.240.542.1107
Global Science & Technology
7855 Walker Drive, Suite 200
Greenbelt, Maryland 20770


-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mailman.ccsds.org/pipermail/css-csts/attachments/20110329/b101bd3d/attachment.htm