[Sea-sa] Revised slides from 14 Nov RASDS++ working meeting

Wed Nov 30 03:21:13 UTC 2022

Here are some possible extensions and revisions to the diagrams, pending discussion.

chart 15 says that a function produces information and has logical structure,

and I think it says that information has data which consumes the logical structure of a function.

Change:

The "consumes" line should go between the headers of function and information instead of between logical structure and data,

and the line should point from function to information.

chart 21 shows type as an attribute of a functional object.

Type is also an attribute of an information object in chart 28.

Type is also an attribute of an operations object in chart 70, using the enumeration {task, activity, process}

Connection type is an attribute of a physical object in chart 88, using the enumeration {flow, energetic, structural, ...}

These types are different.

The information object type specifies its content, and it may additionally specify a layout for concrete objects.

The functional object type names an abstract collection of interfaces and behaviors which use information objects,

as if in an object-oriented class.  See chart 120 for UML class, which is what I mean by "object oriented class".

Change:

Rename "type" in chart 21 to "class", which agrees with the object-oriented concept.

Rename "type" in chart 70 to "operation type", and make a correspondence to function "class" attribute in the texts.

Rename "connection type" in chart 88 to "connection mechanism", to avoid confusion with information types.

chart 22 shows a function producing or consuming data.

This correspondence is present in chart 29 for information objects, but it isn't clear that it corresponds to the "data",

unless "data" in chart 22 has the reverse correspondence.

Change:

There should be a correspondence to information objects for "data" in chart 22.

chart 28 shows information objects without service interfaces.  This implies that functional objects would provide those

interfaces, and information objects describe the layout and content of information objects that pass through those interfaces.

The two viewpoints together provide object oriented models.

Change:

In the text for functional objects, identify this connection to object oriented concepts.

In the text for information objects, say that behavior, interfaces, and management are provided by functional objects.

The access/security concern for information objects will be enforced by functional objects, so perhaps that concern

should move from information objects (chart 28) to functional objects (chart 21).

chart 29 says that information has structure.

chart 31 says that information may have a representation which may have a structure and/or an encoding.

Change:

These two charts together suggest that chart 29 should say "representation" in place of "structure".

chart 30 shows inheritance, composition, aggregation, association, and directed association relationships.

chart 29 shows only the composition relationship.

Change:

Change the composition line in chart 29 to an association line, and change the label on the line to "relationship".

chart 44 does not mention SDU's.

chart 45 mentions SDU's.

Change:

Mention SDU's on chart 44.

chart 65 shows a service processing data.

The correspondence for data is present in chart 22 for functions and information objects,

but it isn't clear that the data corresponds to the data in chart 65.

Change:

There could be a correspondence to information objects for both "data" and "function" in chart 65.

There could be a correspondence to data processed by a service in chart 22.

chart 71 shows a system element accessing data.

It isn't clear that the data corresponds to the data in chart 65.

Change:

Add a correspondence between operational service in chart 71 to service in chart 65.

From: Shames, Peter M (US 312B) <peter.m.shames at jpl.nasa.gov> 
Sent: Tuesday, November 15, 2022 5:33 PM
To: Ramon Krosley <r.krosley at andropogon.org>; Radulescu, Costin (US 9300) <cradule at jpl.nasa.gov>; Fred Slane <frederick.slane at gmail.com>; Christian Stangl <Christian.Stangl at dlr.de>; Davarian, Faramaz (US 9700) <faramaz.davarian at jpl.nasa.gov>
Cc: SEA-SA <sea-sa at mailman.ccsds.org>
Subject: Revised slides from 14 Nov RASDS++ working meeting

Dear RASDS++ team,

Attached are the revised set of slides from the 14 Nov 22 working meeting.  During the meeting we discussed the following aspects:

1.	Aligning with the top level ontology (now pg 15, and updating it as needed)
2.	Adopting the approach of having an “object” figure, and “ontology” figure, and a “representation” figure for each viewpoint, in that order
3.	Adding a prototype functional ontology, along with correspondences, to the intro section of the document providing a visualization for the description of correspondence
4.	Make judicious use of other correspondence relationships in viewpoint ontology views, for context and connections
5.	Agree to use a modified set of UML relationships for the ontology diagrams, introduce these UML relationships early in the document.
6.	Recognize that the ontology diagrams are essentially instances of Information Viewpoint diagrams, adopt that style and document it
7.	Clean up use of terms and make them consistent across viewpoints
8.	Agree to adopt a single, common, representation across all of the ontology diagrams (I chose to use the “RASDS++ standard” instead of alternatives like OML, since we already had most of these in that style)
9.	We can revisit this last item if there is strong feeling of the value of OML and if we can get the OML diagrams to use the adopted information object representation (rounded rectangles)

The attached package has a first cut at all of these changes.  This included creating several new (missing) “object” and “ontology” figures and redrawing two of the existing ones to make them all align.  I also added a new UML relationship figure as slide 11, along with the definitions of these terms “aggregation”, “composition”, “inheritance”, etc.  All views now have all three “standard” diagrams, but these need review by the team.  They also include a limited number of context correspondences, drawn using a lighter color and dashed lines.

A number of “in process” slides were left in place as a reminder of how we got to where we are.  Some of these contain definitions that we will need to clean up, normalize, and incorporate in the body of the text.  The older ontology charts that Ramon and Costin provided were left in place as a cross-check and a reminder of what we had.  I think we are still missing the latest of Costin’s Operational viewpoint updates relating to temporal aspects.

Retaining some of the charts, especially pgs 59-60 is an open question.  As is inclusion of pgs 71-72.  The integration of temporal aspects of operations, pg 73 and elsewhere is also up for discussion as is much of 80-85, a sort of crude operational animation.

Please review and provide any updates, corrections, suggestions, before the next meeting.

Thanks, Peter

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.ccsds.org/pipermail/sea-sa/attachments/20221129/a447bbc4/attachment.htm>