Skip to main content


Eclipse Community Forums
Forum Search:

Search      Help    Register    Login    Home
Home » Modeling » Papyrus » SysML SubModel problems(SysML submodels do not allow SysML diagrams to be created when opened directly.)
SysML SubModel problems [message #1816617] Fri, 01 November 2019 14:46 Go to next message
Philip Spiby is currently offline Philip SpibyFriend
Messages: 8
Registered: October 2014
Junior Member
In a large international project we are wanting to separate a model into a number of separate sub models to allow us to develop in a collaborative environment.

In a simple test using 2019-9 release of Eclipse and Papyrus 4.5 and SysML 1.4 (1.3.2).

1. Create a new Papyrus project
2. Select SysML 1.4 as the context
3. Do not select any initial diagram
4. Create a new child package under the model
5. Create a block definition diagram in the new package
6. Create a block inside the new package and show it on the BDD
7. Create a Parametric diagram inside the block
8. Select the child package and create a submodel selecting the independent submodel resource
9. Save files

When the new submodel is then opened directly it contains the block definition diagram but marked as a class diagram, and the parametric diagram is marked as a Composite structure diagram. The user can create new SysML children and relationships within the submodel, but unable to create any new SysML diagrams.

Is it something we are doing wrong, or is there a problem with Papyrus SysML and submodels?
Re: SysML SubModel problems [message #1816621 is a reply to message #1816617] Fri, 01 November 2019 16:38 Go to previous messageGo to next message
Carsten Pitz is currently offline Carsten PitzFriend
Messages: 477
Registered: May 2015
Location: Germany
Senior Member
Hi Philip,

splitted SysML14 models with Papyrus work perfectly for me.

But I use model elements instead package elements to explicitly indicate these are independent models. But that should not matter at all.

I created a small demo on the configuration (SysML14 1.3.2, on Papyrus 4.5) you mentioned.

The model is structured according to OMG UAF (Unified Architecture Framework) which implements ISO/IEC/IEEE 42010:2011, Systems and software engineering Architecture Description.

Please feel free to import that demo via Import --> Existing Projects into Workspace.

Best regards
Carsten


EDIT typo

[Updated on: Fri, 01 November 2019 16:40]

Report message to a moderator

Re: SysML SubModel problems [message #1816866 is a reply to message #1816621] Fri, 08 November 2019 17:37 Go to previous messageGo to next message
Philip Spiby is currently offline Philip SpibyFriend
Messages: 8
Registered: October 2014
Junior Member
Hi Carsten,

Thanks for your response.
I have downloaded and tried your project on two different systems (just to see if there is a problem with my installation).
On both systems I have the same effect.

If I open MotionController UseCaseDiagram from under Splitted SysML14 Model Demo I can see under the general tab that this is a SysML 1.4 :: SysML 1.4 Use Case Diagram I can also add a SysML 1.4 Activity Diagram into the same parent model (Motion Controller Operational Model).
However if I open the split package Motion_Controller_Operational_model and look at the same UseCaseDiagram I see under the general tab it is of type UML :: Use Case Diagram and if I try to add a diagram only the UML variants of the diagrams are available i.e. UML Activity Diagram and not SysML 1.4 Activity Diagram.

So I think you are experiencing the same effect that I was experiencing.

Could you confirm that when you open the submodel directly from the Project explorer you still have access to the SysML functionality or if this is only available if you navigate to it from the main model.

Best regards
Phil
Re: SysML SubModel problems [message #1816912 is a reply to message #1816866] Mon, 11 November 2019 17:13 Go to previous messageGo to next message
Carsten Pitz is currently offline Carsten PitzFriend
Messages: 477
Registered: May 2015
Location: Germany
Senior Member
Hi Philip,

you can avoid the bahavior by providing an own project for each model and reference between models using the package import mechanism. I attached an example to show this looks like.

I prefer this way, because it makes each model a configuration item, I assume, as AP233 guy (aren't you) you also prefer this way ;-)

BTW, I currently prepare a paper on modeling in large I plan to publish on https://modeling-languages.com/. It ressembles my experiences of the last 22+ years. Including on how to deal with 60 million requirements. I try to explain thing as brief as possible. But nevertheless currently it is 18 pages, with 13 empty chapters still waiting for content. It gets a tour de force ;-)

/Carsten
  • Attachment: Example.zip
    (Size: 10.23KB, Downloaded 76 times)
Re: SysML SubModel problems [message #1818033 is a reply to message #1816912] Wed, 04 December 2019 14:20 Go to previous messageGo to next message
Philip Spiby is currently offline Philip SpibyFriend
Messages: 8
Registered: October 2014
Junior Member
Hi Carsten,

Thanks for you reply, sorry I haven't got back to you until now I have been tied up with other topics for a while (AP239 now!)
I will have a look at your example later today. I agree having configuration data on each separate portion of the model is a good idea, however we are leaving much of that to Git and trying to generate models as they would be if developed directly in Papyrus.
I still believe independent subModels should be editable in the same way if they are edited directly or through the parent model.

RE: AP233 - yep that is me!
We are using SysML to develop the new suite of STEP Application Protocols (with significant input from myself ;-)
The currently selected tool is MagicDraw but I am pushing to adopt Canonical XMI for the models and Canonical DI for the diagrams to ensure the published standards are tool independent (a good requirement of ISO!)
As a consequence I am investigating XMI/DI conversion into Papyrus to the point where you cannot tell if the model was generated directly in Papyrus or not.
This involves a lot of playing around with Papyrus, hence my questions.

Best regards
Phil
Re: SysML SubModel problems [message #1818044 is a reply to message #1818033] Wed, 04 December 2019 17:15 Go to previous messageGo to next message
Carsten Pitz is currently offline Carsten PitzFriend
Messages: 477
Registered: May 2015
Location: Germany
Senior Member
Hi Philip,

AP239.org makes me laugh. When I started to design a S1000D based documentation system 2 decades ago, the S1000D already was version 1.6 and the first drafts of the XML based S1000D version 2.x were out. And AP239.org states S1000D as tomorrow specification :-)

BTW, the IETP-X is still in use ;-)

/Carsten
duplicate [message #1818045 is a reply to message #1818033] Wed, 04 December 2019 17:15 Go to previous messageGo to next message
Carsten Pitz is currently offline Carsten PitzFriend
Messages: 477
Registered: May 2015
Location: Germany
Senior Member
duplicate

[Updated on: Wed, 04 December 2019 17:15]

Report message to a moderator

Re: duplicate [message #1818051 is a reply to message #1818045] Wed, 04 December 2019 19:21 Go to previous messageGo to next message
Philip Spiby is currently offline Philip SpibyFriend
Messages: 8
Registered: October 2014
Junior Member
I have little to do with the marketing of AP239, that is up to our AIA/ASD colleagues.
I am working on the modelling at the core of AP239 and AP242. The latest plans are to look at replacing the ASD Common Model with AP239 edition 3 next year so all the S spec's will map to AP239 (hopefully in a consistent manner).

We are using SysML parametric diagrams to specify the mappings (much better than developing yet another modelling language that STEP is famous for!)

Best regards
Phil
Re: duplicate [message #1818149 is a reply to message #1818051] Sun, 08 December 2019 17:07 Go to previous message
Carsten Pitz is currently offline Carsten PitzFriend
Messages: 477
Registered: May 2015
Location: Germany
Senior Member
Hi Philip,
the paper is published.
https://modeling-languages.com/three-principles-model-complex-systems/
/Carsten
Previous Topic:from CSV to UML use case diagram
Next Topic:SysML Item Flow - Is it possible to hide the "flow" stereotype
Goto Forum:
  


Current Time: Thu Mar 28 17:50:19 GMT 2024

Powered by FUDForum. Page generated in 0.02901 seconds
.:: Contact :: Home ::.

Powered by: FUDforum 3.0.2.
Copyright ©2001-2010 FUDforum Bulletin Board Software

Back to the top