FW: [Ccsds-omg-liaison] Fwd: Spacecraft Operations Language Metamode revised submission available

Tue Feb 22 12:35:20 EST 2011

FYI -

Nick is one of our local UML & modeling experts. He has some cogent and troubling feedback on this OMG effort.   Note also his comment that there is no reference to CCSDS in this work, again.

Peter

From: Nicolas F Rouquette <Nicolas.F.Rouquette at jpl.nasa.gov<mailto:Nicolas.F.Rouquette at jpl.nasa.gov>>
Date: Tue, 22 Feb 2011 07:48:46 -0800
To: Brad Kizzort <bkizzort at harris.com<mailto:bkizzort at harris.com>>
Cc: Andrew Watson <andrew at omg.org<mailto:andrew at omg.org>>
Subject: [Ccsds-omg-liaison] Fwd: Spacecraft Operations Language Metamode revised submission available

Andrew suggested I forward my review to the submitter & to the space mailing list. - Nicolas.

Begin forwarded message:

From: "Rouquette, Nicolas F (313K)" <nicolas.f.rouquette at jpl.nasa.gov<mailto:nicolas.f.rouquette at jpl.nasa.gov>>
Date: February 21, 2011 11:00:14 PM PST
To: Andrew Watson <andrew at omg.org<mailto:andrew at omg.org>>, "Shames, Peter M (313B)" <peter.m.shames at jpl.nasa.gov<mailto:peter.m.shames at jpl.nasa.gov>>
Cc: "Jenkins, J Steven (3101)" <j.s.jenkins at jpl.nasa.gov<mailto:j.s.jenkins at jpl.nasa.gov>>, Sanford Friedenthal <safriedenthal at gmail.com<mailto:safriedenthal at gmail.com>>, Pete Adaptive <pete.rivett at adaptive.com<mailto:pete.rivett at adaptive.com>>, Juergen Boldt <juergen at omg.org<mailto:juergen at omg.org>>, Ed Seidewitz <ed-s at modeldriven.com<mailto:ed-s at modeldriven.com>>
Subject: Re: Spacecraft Operations Language Metamode revised submission available

Andrew, Peter,

I have reviewed the space/2011-02-01 specification and found serious issues from NASA's perspective that fall in 4 categories:

1) SMSC-related problems
2) Specification architecture-related problems
3) systems engineering
4) space communications standards.

Details follow.

1) SMSC-related problems:

a) no bookmarks and no hyperlinks makes navigating the document fastidious.
b) some constructs are referenced but not defined; e.g. AbsoluteTime.

2) Specification document/architecture-related problems:

c) The semantics of elements is given in fragments; e.g., the semantics of ExecutableNode is in 3 different places:

6.2.7 ControlFlow, p. 16, see 6.2.7.5
6.2.10 ExecutableNode, p. 17
6.2.14 HandledExceptionRegion

d) UML/FUML vs. SOLM

SOLM 6.2.7.5 Control Flow semantics states:

Each ExecutableNode or ControlNode in a control flow completes its action before control is passed
through a ControlFlow to the next ExecutableNode or ControlNode in the chain.

SOLM 6.2 Activities is very misleading:

SOLM closely follows the Activity modeling in UML and Foundational UML,
but there are some simplifications to the meta-model,
since SOLM is not intended as a general purpose software metamodel.

SOLM borrows UML's abstract syntax vocabulary but with a different semantics.
The semantics for UML activities defined in FUML is based on the concept of activation; see FUML clause 8.2.2.1
The semantics in SOLM for ExecutableNode is based on completion; see SOLM clause 6.2.75.
These are two very different semantics for activities that are not equivalent to each other.
If anything SOLM's completion semantics could be described as a specialization of BPMN2's completion semantics -- see BPMN2, clause 13.2.2

Based on the above analysis, I believe that the compliance statement for RFP requirement 6.5.4 on p. 44 is false:

SOLM is a MOF Meta-model that closely follows the meta-model underneath the UML Activity Diagram, with domain specific actions and nodes.

3) Systems engineering perspective

e) The model of Time (6.3.15) , RelativeTime (SOLM 6.3.10) and AbsoluteTime (6.3.1) is inadequate for the space mission domain.

This specification document is clearly written from the perspective of spacecraft operating in Earth orbit where roundtrip transmission delays are stable and short (between 0.05 seconds for Low Earth Orbiting (LEO) satellites at 500-to-1,000 miles altitude to 0.25 seconds for Geostationary Earth Orbiting (GEO) satellites at 22,238 miles altitude). In these conditions, is possible to simplify the spacecraft design to operate according to a clock synchronized with a clock reference on Earth.
This strategy allows using simple protocols to maintain accurate knowledge in the ground system of the parameter values on the spacecraft as shown in figure 3, p. 7.

The lengthy round-trip transmission delays characteristic of deep-space missions prevent simple state synchronization protocols as the one shown in figure 3, p 7.
For example, spacecraft operating in Mars orbit have roundtrip transmission delays ranging between 23 to 40 minutes due to the changes in Mars/Earth distance.
For deep-space missions, it becomes important to distinguish time at different physical loci; such as an Earth clock vs. a spacecraft clock. It is also important to avoid designing mission operations that require lengthy exchanges between these clocks. Defining multiple clocks goes hand-in-hand with defining multiple execution loci (see FUML clause 8.2).

f) Execution loci.

The Foundational UML specification explicitly defined the concept of execution locus; i.e., the place where actions are executed by an executive and where the effects of these actions can be observed. Although SOLM explicitly refers to FUML, it uses a different terminology, that of environments of different kinds:
- A SOLM execution environment (clauses 1.1, 1.2, 2, 6)
- A SOLM executive (clause 5, Glossary, p. 6)
- ProcedureEnvironment (clause 6.3.11)

Based on the above analysis, I believe that the compliance statement for RFP requirement 6.5.2 on p. 44 is inadmissible.

4) space communication standards.

I defer to my colleague, Peter Shames, who is an expert in CCSDS standards to give you a critique of the SOLM submission.

- Nicolas.

On Feb 21, 2011, at 4:56 PM, Sanford Friedenthal wrote:

Interesting.

From: Juergen Boldt [mailto:juergen at omg.org]
Sent: Monday, February 21, 2011 6:27 PM
To: Recipient list suppressed:
Subject: Spacecraft Operations Language Metamode revised submission available
Importance: High

All,

a revised submission to the Spacecraft Operations Language Metamodel RFP has been posted to the OMG document as space/2011-02-01 at
http://www.omg.org/cgi-bin/doc?space/2011-02-01

The submission is accompanied by XMI space/2011-02-02, EAP space/2011-02-03 as well as the inventory file which has the document number space/2011-02-04.
The submission has been linked to the work in progress page located at
http://www.omg.org/techprocess/meetings/schedule/Operations_Procedure_MM_RFP.html

The submitter requested a vote to vote for the Arlington meeting, and thuis the submission has been placed onto the AB and DTC agendas for endorsement/adoptionj vote

Best regards
Juergen Boldt
Director, Member Services
140 Kendrick Street, Building A Suite 300
Needham, MA 02494 USA

Tel:  781 444 0404 x 132
fax:  781 444 0320

www.omg.org<http://www.omg.org/>

[http://www.omg.org/images/signature-2.gif]<http://www.omg.org/signature.htm>

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mailman.ccsds.org/pipermail/ccsds-omg-liaison/attachments/20110222/ec18f28b/attachment.htm