[Ccsds-omg-liaison] FW: Initial AB Review of Space Ops Language Metamodel Submission : space/11-02-01..04

[Hooke, Adrian J (9000)]

From: Sridhar Iyengar [mailto:siyengar at us.ibm.com]
Sent: Friday, March 11, 2011 12:23 AM
To: ab at omg.org; space at omg.org; bkizzort at harris.com
Subject: Initial AB Review of Space Ops Language Metamodel Submission : space/11-02-01..04

Documents:
-----------------
space 11/02-01 : Revised Submission
space 11/02-02 : XMI files (XMI for SOLM Metamodel)    :  This is not an EMOF or CMOF metamodel  file.  This is a UML model that was treated 'as a metamodel'.
space 11/02-03 : Convenience files in Zip archive (EAP version of metamodel)  : I dont have EAP - so cant verify the file. I will assume it represents the UML model representing the metamodel
space 11/02-04 : Inventory file

Comments
---------------------------
Nice revision history. Surprised this has been going on since 2005!

Spec looks comprehensive.  Nice to see both SpacePython and CCL mappings.

The XMI file 11/02-02 is NOT a MOF model - It is a UML model that can be used to represent the metamodel.  We need a normative MOF metamodel file.

Requirements traceability is complete.

Inventory file is complete

Page 2 : Last para : Why define SOLM as another metamodel for process modeling if that is the primary usage? Why not just use BPMN or UML Activity Diagrams? Is the something specific in spacecraft operations that is not easily handled in BPMN or UML?  If so , it would be good to call this out.  Section 6.0 and 6.1 have some of this rationale.  Are you still able to use UML tools to document your activities graphically and use the tool to generate the scripts needed?  I am assuming you already tried the 'UML Profile approach' and chose to define a metamodel based on that experience (I looked at the spec history).  Having an errata to the spec with a clear justification would be quite useful. I expect this to be a discussion point in our review.

Page 5 : Note that newer versions of UML, MOF and XMI have been developed than the ones referenced.  These need to be updated unless there is a specific reason for sticking with the earlier versions

Page 6 : I see that you have added a new definition of meta model.  OMG already has a standard definition in its glossary and in the RFP template.


Page  31. Section 6.4.0

I see quite a few uses of 'composite associations' in the metamodel - please make sure this is your intention.









Regards,
Sridhar Iyengar
E-Mail :  siyengar at us,ibm.com Phone : +1-919-593-4068
IBM Distinguished Engineer, TJ Watson Research Center
Assistant : Felicia Medley, feliciam at us.ibm.com 1-914-784-7068
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mailman.ccsds.org/pipermail/ccsds-omg-liaison/attachments/20110311/27db2ce2/attachment.html