|
|
|
Re: Any Interest in UAF [message #1847187 is a reply to message #1847184] |
Mon, 18 October 2021 15:35 |
|
Hi Rod,
I am about to roll on such a profile for Papyrus, if I have a longer project free period or a 3 days/week project. But I was fully booked the last years.
Is your interest UAF directly or the NAFv4?
Is a profile sufficient or shall guidance be provided?
/Carsten
|
|
|
|
Re: Any Interest in UAF [message #1847258 is a reply to message #1847222] |
Thu, 21 October 2021 11:02 |
|
Hi Rod,
NAFv4 is the NATO's adoption of the UAF. BTW, ADMBw is the German Armed Forces' adoption of the NAFv4.
OK, having a NATO frigate as system to design I would use NAFv4 ;-)
For example NAFv4
does not use all views UAF defines
/Carsten
|
|
|
|
Re: Any Interest in UAF [message #1847266 is a reply to message #1847263] |
Thu, 21 October 2021 13:52 |
|
Hi Carsten, Rod and anyone else that might be interested
I made an attempt to prepare a profile for Papyrus based on the official OMG-profile for UAF.
References to OMG SysML 1.6 profile has been changed to the Papyrus SysML 1.6 profile provided in the SysML 1.6 extension.
References to primitive types has been changed to references to UML2 primitive types
But I run into several issues...
These issues had to be taken care of before the profile could be released and be applied to a model:
Papyrus seems have problem applying stereotypes defined within a profile package with spaces in its name. Spaces in package names have therfore been removed.
<<UAF::Summary and Overview::View>> defines a "viewpoint" property which conflicts with the inherited "viewPoint" properties defined in <<SysML::ModelELements::View>>. I have therefore removed the UAF::Summary and Overview::View viewpoint property.
<<UAF::Summary and Overview::ViewPoint>> defines concern and method properties which conflicts with the inherited "concern" and "method" properties defined in <<SysML::ModelELements::ViewPoint>>. I have therefore removed the <<UAF::Summary and Overview::ViewPoint>> "concern" and "method" properties.
<<UAF::Projects::Roadmap::ActualProjectMilestone>> has an endDate with upper limit 0 which is not allowed. This seems like a bug in the OMG profile. I changed the upper limit to 1.
The multiplicity of all properties refering to UML metaclasses (base_XXX) have been changed from 1 to 0..1.
I have attached the migrated and released profile with those adjustments to this post.
But even after the profile was released, some more issues appear when applying the profile to a model.
It seems like there are conflicts when UAF stereotypes extends UML metaclasses and in the same time inherits from SysML stereotypes extending the same metaclasses
The following type of messages are shown when loading a model with the UAF-profile applied containg a class with stereotype <<System>> applied:
org.xml.sax.SAXParseException; systemId: platform:/resource/SysML%20Test/T1.uml; lineNumber: 18; columnNumber: 300; Attribute "base_Class" was already specified for element "UAF.Resources.Taxonomy:System".
When you look in the model´s UML-file you will see that "base_Class" is specified twice:
<UAF.Resources.Taxonomy:System xmi:id="_6INq8DJyEeyMzdFzXDF1Cw" URI="ddwdwwd" base_Element="_5H9u4DJyEeyMzdFzXDF1Cw" base_Class="_5H9u4DJyEeyMzdFzXDF1Cw" base_Class="_5H9u4DJyEeyMzdFzXDF1Cw" base_SubjectOfForecast_Class="_5H9u4DJyEeyMzdFzXDF1Cw" base_Architecture_Class="_5H9u4DJyEeyMzdFzXDF1Cw"/>
I assume that the solution of this issue is to go through all UAF stereotypes and remove redundant extensions to UML metaclasses, or do anyone have any better idea?
Thanks,
Thomas
Thomas Wiman
MetaModelAgent Product Manager
|
|
|
|
|
|
|
Re: Any Interest in UAF [message #1847435 is a reply to message #1847419] |
Wed, 27 October 2021 06:56 |
|
Thank you Carsten, I blush on behalf of Sweden. ;-)
In fact, I had problems myself applying the profile in my environment (Papyrus 2021-09 with SysML 1.6 installed). That's what I mentioned in the previous post.
But when I now try again, it works fine.
I have no knowledge at all in UAF. The profile is based on SysML, so that profile must be installed through the SysML extension to work. But is the idea that in UAF modeling one should use the concepts (stereotypes) in SysML as well? If so, the SysML profile needs to be applied explicitly.
Otherwise, I suggest that in Papyrus using the New Model wizard first create a standard UML model (with architectural context UML) and then apply the UAF profile (via the profile tab in Property View).
Of course, it would be easier with a customized GUI (Model explorers context menu and diagram palette) to create UAF-specific elements, but it is a much bigger challenge and requires knowledge of how UAF should be applied to be right...
/Thomas
Thomas Wiman
MetaModelAgent Product Manager
|
|
|
Re: Any Interest in UAF [message #1847451 is a reply to message #1847435] |
Wed, 27 October 2021 15:07 |
|
In the absence of a custom UAF User Interface, I have generated an Excel sheet, which I attach to this post, which shows which UML elements the respective UAF stereotypes can be applied to.
Hope it can be of some help.
/Thomas
Thomas Wiman
MetaModelAgent Product Manager
|
|
|
|
|
|
|
Re: Any Interest in UAF [message #1863756 is a reply to message #1863714] |
Sun, 25 February 2024 12:36 |
|
Hi Rod,
Even old threads are allowed to get active again...
The result of applying the styleeshet to a model or to a diagram is that all elements having UAF stereotypes applied will be displayed using the correct UAF color scheme with all compartments hidden.
See attached image.
/Thomas
Thomas Wiman
MetaModelAgent Product Manager
|
|
|
Powered by
FUDForum. Page generated in 0.06696 seconds