[New Model]: Pi10 new model IdBasedComment

[nhaenis]

Description

New model for an ID based data exchange of comments belonging to a UUID / entity.

Closes #308

MS2 Criteria

(to be filled out by PR reviewer)

• the model validates with the SAMM SDS SDK in the version specified in the Readme.md of this repository by the time of the MS2 check (e.g., 'java -jar samm-cli.jar aspect <path-to-aspect-model> validate ). The SAMM CLI is available here and in GitHub
• use Camel-Case (e.g., "MyModelElement" or "TimeDifferenceGmtId", when in doubt follow https://google.github.io/styleguide/javaguide.html#s5.3-camel-case)
• the identifiers for all model elements start with a capital letter except for properties
• the identifier for properties starts with a small letter
• all model elements at least contain the fields "preferred name" and "description" in English language. The description must be comprehensible. It is not required to write full sentences but style should be consistent over the whole model
• Property and the referenced Characteristic should not have the same name
• the versioning in the URN follows semantic versioning, where minor version bumps are backwards compatible and major version bumps are not backwards compatible.
• use abbreviations only when necessary and if these are sufficiently common
• avoid redundant prefixes in property names (consider adding properties to an enclosing Entity or even adapt the namespace of the model elements, e.g., instead of having two properties DismantlerId and DismantlerName use an Entity Dismantler with the properties name and id or use a URN like io.catenax.dismantler:0.0.1)
• fields preferredName and description are not the same
• preferredName should be human readable and follow normal orthography (e.g., no camel case but normal word separation)
• name of aspect is singular except if it only has one property which is a Collection, List or Set. In theses cases, the aspect name is plural.
• units are referenced from the SAMM unit catalog whenever possible
• use constraints to make known constraints from the use case explicit in the aspect model
• when relying on external standards, they are referenced through a "see" element
• all properties with an simple type have an example value
• metadata.json exists with status "release"
• generated json schema validates against example json payload
• file RELEASE_NOTES.md exists and contains entries for proposed model changes
• all contributors to this model are mentioned in copyright header of model file

MS3 Criteria

(to be filled out by semantic modeling team before merge to main-branch)

• All required reviewers have approved this PR (see reviewers section)
• The new aspect (version) will be implemented by at least one data provider
• The new aspect (version) will be consumed by at least one data consumer
• There exists valid test data
• In case of a new (incompatible) major version to an existing version, a migration strategy has been developed
• The model has at least version '1.0.0'
• If a previous model exists, model deprecation has been checked for previous model
• The release date in the Release Note is set to the date of the MS3 approval

[github-actions]

Validation Report for io.catenax.id_based_comment/1.0.0/IdBasedComment.ttl

Input model is valid

[LukasHeimann]

This model is lacking customer and supplier. However, they are crucial for this model to be sound. Let me elaborate:

A customer-supplier relationship is a directed one: One party is customer, one is supplier. Comments, as well as the Material Demands and Capacity Groups they refer to, have IDs. These IDs are, in their description, explicitly defined to only be unique "within the business relationship between a customer and its supplier". Hence, to uniquely resolve a comment by its commentId, as well as a Material Demand or Capacity Group by an $(objectId, objectType)$ tuple, you need to also know the supplier and the customer.

It is not possible to infer this information from the message header, as the message header only contains information about the sender and the receiver of the message. However, as both customer and supplier can send (create/update/delete) comments, it is not possible to know if the sender of the message is the customer or the supplier. Thus, it is not possible to derive the correct customer-supplier relationship. Hence, we can't uniquely address the comment, material demand or capacity group.

Example: As per our standard, it is valid that two companies $A$ and $B$ are both supplier and customer to each other -- so there are two business partner relationships $(customer, supplier)$: $(A, B)$ and $(B, A)$. In both relationships, a material demand can be created with the same UUID. $B$ creates a Material Demand for $A$ with ID 123, and $A$ creates a Material Demand for $B$ with ID 123. The standard, today, forces us to not throw an error but treat those as separate Material Demands (because the ID must be treated as unique only within the relationship, not globally). Now, A posts a comment on objectId 123. Which Material Demand should receive the comment? A surely is the sender and B the receiver of the message, but from that you can't derive whether the business partner relationship $(A, B)$ or $(B, A)$ is targeted, because $A$ could have commented on the Material Demand where it is supplier, or on the one where it is customer.

I know this is an edge case, but we allow (and must allow) sender-controlled IDs, and the other standards define them as not globally unique (allowing us to fail when we receive a second material demand (or comment/capacity group) with the same UUID but a different business partner relationship), but uniquely valid only within the business partner relationship (forcing us to accept them). Thus, the standards must also deal with this edge case.

[LukasHeimann]

properties were added as discussed in the architecture meeting today

[github-actions]

Validation Report for io.catenax.id_based_comment/1.0.0/IdBasedComment.ttl

Input model is valid

[agg3fe]

You can remove completely the 'gen' folder as this will be generated at run time automatically.

[agg3fe]

you can make use of shared aspect model (io.catenax.shared.uuid:1.0.0) for UUID.
you can refer this https://github.com/eclipse-tractusx/sldt-semantic-models/blob/a07bdd5b1515767d4f8d07deb47cf2a4aa728bef/io.catenax.quality_message_content/1.0.0/QualityMessageContent.ttl#L35C9-L35C17

[agg3fe]

Here also please check if you can make use of shared aspect model https://github.com/eclipse-tractusx/sldt-semantic-models/blob/a07bdd5b1515767d4f8d07deb47cf2a4aa728bef/io.catenax.shared.business_partner_number/1.0.0/BusinessPartnerNumber.ttl

[LukasHeimann]

I love it, this would make sense for the other models of DCM as well!

[agg3fe]

would be great to provide a bit more elaborated description here.

[agg3fe]

please check my comments

[github-actions]

Validation Report for io.catenax.id_based_comment/1.0.0/IdBasedComment.ttl

Input model is valid

[nhaenis]

@agg3fe & @LukasHeimann: model uses shared BPN's + UUID aspect now, fixed description for the text. Ready for review

[agg3fe]

typo belongig

[agg3fe]

@nhaenis could you please correct a small typo mentioned in my PR comment. then we can do the MS2 today.
Also please change the date in the release notes. you can put today's date in case we do the MS3 also today.

[github-actions]

Validation Report for io.catenax.id_based_comment/1.0.0/IdBasedComment.ttl

Input model is valid

[nhaenis]

@agg3fe the picture in the HTML in gen folder is not generated correctly. Is there already a bug in SAMM CLI opened for a fix of the visual representations of the aspects when another aspect is added as ext-ref? Locally i have the same issue and need to comment out the lines of the models UUID and BPN (lines 24 to 29 in both BPN aspect and UUID aspect) in order to generate it properly

[agg3fe]

@agg3fe the picture in the HTML in gen folder is not generated correctly. Is there already a bug in SAMM CLI opened for a fix of the visual representations of the aspects when another aspect is added as ext-ref? Locally i have the same issue and need to comment out the lines of the models UUID and BPN (lines 24 to 29 in both BPN aspect and UUID aspect) in order to generate it properly

@nhaenis I think there are some issues going on with the SDK and they are working on it. I will have a look into it.