[xDFB] Display Function's owner

Hello everyone,
I’m wondering why Dataflow diagrams don’t display the owner of each Function. I understand that there is green / blue colour management, but when I design a scenario with an xDFB, it would be very useful to express statements such as “Actor X requests something” instead of “Someone outside the system requests something and if you are reading this diagram outside Capella (thus with no Semantic Browser) you’re screwed if you need to know who it is”.
Did I miss a feature?
Best regards,
Barry Allen

Hello Barry,
if you are interested in functions / functionalities assingned to components/actors why don’t you use a [xAB] diagram in this case?

You could also manually put notes with the note tool (in the right Palette) and order the note below the Function involvement in the xDFB diagram (Rt click > format > order > send to back) with the name of the component/actor. It wont update automatically, and it wont link to a component if using the standard HTML export tool, but its a start for some additional custom context. :slight_smile: Otherwise a xAB diagram contextual to the elements you are interest may be helpful.
Brandon

Quote:
if you are interested in functions / functionalities assingned to components/actors why don’t you use a [xAB] diagram in this case?
Because I’m not interested in an architecture diagram, but in describing a behaviour. I would like to “replicate” a SysML Activity Diagram.
I think that a diagram that reads “this function is inside the system, that one is outside” has no added value. If a function is outside, then say where it is exactly, especially if you point its externality out by painting it in blue.
Quote:
You could also manually put notes with the note tool (in the right Palette) and order the note below the Function involvement in the xDFB diagram (Rt click > format > order > send to back) with the name of the component/actor.
This would be overly complex. It is the same as if you said to me “you could also perform your modelling with a pen and paper”. I could, but I want things to be easy and automated, that’s why I expect from a tool.
The issue we are discussing in this topic is almost a philosophical one. I definitely don’t understand the Arcadia process: you have goals that the system must help achieve, but these goals are expressed by no one. In order to describe how something is performed, you model “Functions A and B must be realised” but you don’t say who actually executes these functions. Very strange.

Hi Charles,
Regarding “replicating” a SysML Activity diagram, you’ll find more information on the underlying topic here
https://www.eclipse.org/capella/arcadia_capella_sysml_tool.html#activity and then following the link to the paper mentioned (mixing control flows and data flows).
And here is a webinar that will give you more information on how to manage control flows with functional chains:
https://www.youtube.com/watch?v=ZqvfSURUpIY
I don’t share your comment on the fact that Arcadia does not enable you to say who is performing a function as functions are allocated to components. Anyway if you’re interested to have a better understanding of Arcadia, I do recommend this book from Jean-Luc Voirin:
https://www.elsevier.com/books/model-based-system-and-architecture-engineering-with-the-arcadia-method/voirin/978-1-78548-169-7 (also available in French)
I hope this helps,
Stephane LACRAMPE
Obeo

Copyright © Eclipse Capella, the Eclipse Capella logo, Eclipse and the Eclipse logo are Trademarks of The Eclipse Foundation.