[Ccsds-omg-liaison] CSRM comments

Pete Rivett pete.rivett at adaptive.com
Thu Oct 3 15:40:46 UTC 2019

  *   The SysML namespace resolves to a 404 page error
     *   xmlns:sysml='http://www.omg.org/spec/SysML/20150709/SysML'

That URI is correct for SysML 1.4. Namespaces are URIs (identifiers) not URLs (Locators) so not required to resolve. The URL is https://www.omg.org/spec/SysML/20150709/SysML.xmi
Though it'd be nice if the URI was redirected to the URL that's a matter for the OMG website under instruction from the SysML RTF (for FIBO we specified that sort of redirect).
In short it's not something the CSRM spec needs to address.

However, having just looked at the file there are additional problems:

  *   There is a top level uml:Model called Data which should not be there. The top level element should be the Package called Cubesat System Reference Model.
  *   It contains diagrams and other xmi:Extensions in proprietary MagicDraw format. That includes the HyperlinkOwner elements that JD mentioned
  *   References to external models do not use the official OMG URL but refer a local mdzip file e.g.
  *                                     <elementImport xmi:type='uml:ElementImport' xmi:id='_19_0_2_3880181_1556034789456_254723_58418'>
                                         <importedElement href='ISO-80000.mdzip#_18_0beta_903028d_1388650189782_109507_116243'>
                                                <xmi:Extension extender='MagicDraw UML 19.0'>
                                                       <referenceExtension referentPath='ISO-80000::ISO80000-9 Physical Chemestry and Molecular Physics::Quantities::amount of substance::amount of substance' referentType='DataType'/>

  *   Many problems such as the diagrams and extensions can be addressed by exporting from MagicDraw using the Clean XMI plugin. See attached email for how to install and use.
  *   Note that, in addition to a clean, standard XMI file, you should also include the original MagicDraw proprietary .mdzip file for future maintenance of the specification so long as it is clearly labeled as Ancillary in the specification Manifest file.

In short I suggest you run the Clean XMI plugin and send me the output and I can clean it up further (e.g. remove the top Data element) then give it to JD again to check.


From: omg_jd at objectsandaspects.com <omg_jd at objectsandaspects.com>
Sent: Thursday, October 3, 2019 7:02 AM
To: space at omg.org
Cc: david.kaslow at gmail.com; 'Graham Bleakley' <graham.bleakley at uk.ibm.com>
Subject: CSRM comments

The time between our meeting last week, and the required delivery date for the revised submission is short.  I have begun to import the INCOSE model into EA under the assumption that I will be able to publish it soon.  Since the time to address issues is short, I have decided to publish these initial comments so we can begin to make the necessary revisions.

The XMI file, space_19_08_04, is not standard XMI.  It has a number of MagicDraw proprietary entries and information.  In addition there are some errors in the standard information.  I suggest we find someone who understands the MagicDraw XMI generation to  assist with the generation of acceptable XMI.
1.       It looks like there are several proprietary extensions used in the model:
xmlns:Architecture_Structures_Profile='http://www.magicdraw.com/schemas/Architecture_Structures_Profile.xmi' xmlns:MD_Customization_for_Requirements__additional_stereotypes='http://www.magicdraw.com/spec/Customization/180/Requirements'
xmlns:Validation_Profile='http://www.magicdraw.com/schemas/Validation_Profile.xmi' xmlns:Dependency_Matrix_Profile='http://www.magicdraw.com/schemas/Dependency_Matrix_Profile.xmi'

  1.  The SysML namespace resolves to a 404 page error

     *   xmlns:sysml='http://www.omg.org/spec/SysML/20150709/SysML'

One of the more prevalent stereotypes is <HyperlinkOwner> which has been applied to most but not all of the packages and elements in the model.  I'm not sure what that is, but I assume it relates to how MagicDraw manages elements in their tool and should not appear in OMG XMI.

The Conference Papers package includes SysML notes that reference interim status presentations.  I don't believe these would be applicable to a CubeSat developer using the reference model and should probably be removed.

It looks like there are duplicate named elements for most if not all of the model elements.  For example there is an actor named 18th Space Control Squadron and there is a class named 18th Space Control Squadron (this may be a Block in the original model).  The class seems to exist solely to be a place to capture the description of the stakeholder in a note property.  I do not like the notion of two elements with the same name.  If the class exists simply to capture the description why can't we do this as part of the actor properties?

MagicDraw proprietary tables have been used to map stakeholders to their concerns.  It does not look like the concerns are model elements, rather they are just text in a table.  A proper model (in my opinion) would have the concerns captured as model elements and a relationship between the concerns and the appropriate stakeholder.  Modeling tools could then choose to present this information as a diagram, a matrix or a table, or anything else that might be appropriate.

J.D. Baker
Sparx Systems Ambassador
Mobile - 760 315-3636


Virus-free. www.avast.com<https://www.avast.com/sig-email?utm_medium=email&utm_source=link&utm_campaign=sig-email&utm_content=emailclient&utm_term=link>

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.ccsds.org/pipermail/ccsds-omg-liaison/attachments/20191003/6ab68957/attachment-0001.html>
-------------- next part --------------
An embedded message was scrubbed...
From: Pete Rivett <pete.rivett at adaptive.com>
Subject: RE: [EXTERNAL] Preliminary AB Review: Command and Control Message Specification (C2MS) 1.0 FTF report
Date: Thu, 13 Jun 2019 01:38:55 +0000
Size: 281245
URL: <http://mailman.ccsds.org/pipermail/ccsds-omg-liaison/attachments/20191003/6ab68957/attachment-0001.mht>

More information about the CCSDS-OMG-Liaison mailing list