1. Changes to the model resulting from the XML schema work by Achim/Larry
Some model issues were detected during the XML schema work by Larry and Achim
List of classes in DDICDILibrary:
The identifier attribute should be added to these classes. Dependent on the implementation, these classes could be instantiated. Then it makes sense that they have an identifier.
Action: Flavio to implement
Attributes on Reference: What are the “required” and “semantic” attributes on Reference? They are undocumented.
Decision: Remove “required” and document “semantic”.
Action: Flavio to remove “required”. Arofan and Hilde to document “semantic” as the type of reference that may or may not come from a Controlled Vocabulary.
Cardinality change: Multiplicity of attribute Reference::ddiReference should be 0..1 instead of 1..1.
Decision: Agreed to change the cardinality of reference to 0..1.
Action: Flavio to implement.
NB: New file from Flavio available that includes the agreed changes provided post meeting on 2022-02-16.
entryReference and vocabulary lacks documentation.
Action: Arofan and Hilde to complete this during the field level documentation upgrade.
2. Other issues
It may not be time to provide an XML schema before the release. Larry proposed to provide a snippet for the release, containing Concept, Variable, Codelist.
Larry to continue with the XML and the example.
Arofan and Hilde to continue with the field level documentation upgrade.
Check in with the group next week as needed if more issues may be detected.
Minutes Virtual CDI meeting 2022-01-26
Achim, Arofan, Flavio, Hilde, Dan, Larry, Wendy
- SciDataCon panel plan
Arofan will prepare a draft proposal for a CDI panel for SciDataCon for the group to look at. The CDI as part of an interoperability framework together with other standards like DCAT, http://schema.org , PROV etc.
2. Release Issues:
- Enumeration property questions
Arofan is working on reviewing definitions for Enumerations and had some questions to some of them.
Action: Flavio to look at the definitions for these, and check if changes are needed to the list of items and their definitions.
Action: Flavio to find definitions for these, look at DDI-L (3.3) /possibly Process model in DDI4
Enumerations with no connections:
Action: Flavio will look for enumerations with no connections in the model. This task has been compled.
- Updating the XML and field-level online doc (for in-progress work)
Action: Achim will provide new clickable documentation and new XMI files, based on the EA file from Flavio of 2021-12-21. This task has been completed.
- Plan for incorporating field-level doc
Arofan will continue to revise field level documentation and will send to Hilde for discussion and entry when he has something in hand.
4. Further discussion of moving forward topics/SC doc from this week
Delivery of model to the TC:
Wendy asked Arofan to provide a tentative date for this in the near future.
Document for the Scientific Board Contact:
Arofan has created a document for Simon, the Scientific Board Contact of the CDI group, about the next step priorities of the group.
Arofan reported that CODATA and RDA’s WorldFAIR EU grant proposal has been accepted.
Deadline for abstracts to SciDataCon 2022 is February 14th 2022. Arofan will prepare a proposal for a session on DDI.
2. Status on release
The first step is for Arofan and Hilde will continue the preparation of the field level documentation. The upgrade of the spec is the next step.
3. Further discussion of moving forward topics
Achim is currently working on tools for model to text transformations using Eclipse Acceleo for transform the EA model to canonical xmi. Bitbucket is used an environment for this. A goal is to automate the process as much as possible, however, some manual steps related to the export from EA and the check of output from the transformation are needed.
Wendy pointed out that production flows and tools may vary between the DDI products and that an overview regarding what we have, where things are, how processes are realised etc. would be useful to document for each product.
Achim is currently working to improve the production tool, and will then integrate
the RDF code of Pierre Antoine. He will get back to Pierre Antoine in February regarding this.
-Other focus areas
How other domains look at the CDI model: Flavio will forward a paper he is writing together with colleagues that addresses this.
Implementation guidelines work from Dagstuhl: Arofan will send a draft of the Dagstuhl work to the group for input. Flavio and Wendy are interested to look at this.
Standards mapping: Achim presented a matrix for standards mapping at Dagstuhl. This could form basis for a regular way to document mappings between standards. To have mappings representations in SKOS would allow machine actionability. Wendy pointed out that mappings should preferably be documented at several levels for different purposes, starting from conceptual and going into more detail. Flavio pointed out that we should maintain CDI related mappings.
After the work on the first version of the CDI model and the spec has been finished, the RDF representation would be the next thing on the priorities list, and after that implementation guides and mappings. The current XML representation also needs a review.
Additional syntaxes was also a topic at Dagstuhl. Larry mentioned that we need to think about a method for the mapping of the model to object oriented languages, and not only put focus on the objects. Arofan pointed out that Eclipse based tools can bind to object oriented programming languages (Java, Phyton, C++ etc.). This needs to be further explored and perhaps discussed with modelling experts. Achim pointed out that EA can do this as well, but that it has its own mapping understanding. StatCan also has an approach for general libraries that adress a family of object oriented languages. Like the EA approach this also has some limitations. Wendy pointed out that it would be useful to discuss this topic in a broader group as well. Transformation rules in COGS could for example be explored.
It was agreed that Arofan will organise an effort to write a position paper on the topic later on.
Minutes Virtual CDI meeting 2021-12-15
Achim, Arofan, Dan, Flavio, Hilde, Jay, Larry, Wendy
1. Admin (if any)
- Liaison w/ SB (Simon)
Simon Hodson will be the CDI contact from the Scientific Board. The first point of contact will be that Arofan drafts a document describing goals achieved and future plans of the group.
2. Updates on documentation issues
Arofan and Hilde are working on the field level documentation. All elements have been gone through, as to whether their definitions and other documentation like explanations or examples need to be added or changed. Properties have mostly been edited and is available in a temporary EA file that the group has access to – updates with edits and additions of other components like classes, associations, datatypes and enumerations are forthcoming.
3. Updates on model freeze and generation
Achim has created a new canonical xmi file based on Flavio’s draft frozen file of 2021-11-23, with some minor edits.
Flavio will implement the changes proposed by Achim to the temporary EA file that Hilde is working on and, will produce a new file on which she can continue to edit the field level documentation.
Work on release candidates for xsd is ongoing.
Minutes Virtual CDI meeting 2021-11-10
Achim, Arofan, Dan, Hilde, Jay, Larry
Hilde reported from the meeting of the Scientific Board.
Arofan reported about the tutorials planned for the EDDI Training FAIR.
2. Recommended model freeze and forward game plan (from Friday meeting last week)
It was discussed whether the model is ready to freeze at this point. The topic of whether to keep the patterns in the model was then discussed. It has been agreed earlier not to have the patterns in the model. At this meeting it was agreed to keep the patterns in the model with the reasoning that they can be useful for developers, and that it would not make sense to have one version of the model with patterns and another one without them.
Actions related to the inclusion of patterns:
In relation to the model: Achim to ask Flavio to make pattern classes abstract again. A new version of the model to be provided.
In relation to the documentation:
Arofan/Hilde will review field level documentation for the patterns.
Achim will find out how to make patterns part of the generated documentation.
3. Status and updates
Arofan and Hilde will start to review the filed level documentation. A question that came up at the meeting is how to deal with examples in the documentation. Wendy suggested to have an example file at the product page that could be referenced. Possibilities need to be further explored.
Achim will take care of the XMI production for the release. The production process includes some manual steps. The content is currently stored in Bitbucket.
Achim will liaise with Wendy/TC regarding what the TC needs to make home for in connection with the release (types of material etc.).
Minutes Virtual CDI meeting 2021-10-27
Achim, Arofan, Dan, Flavio, Hilde, Jay,
1. Admin (if any)
DDI-CDI and related usecases, the role of CDI in a core interoperability framework etc. was presented at a SciDataCon, see https://www.scidatacon.org/virtual-2021/sessions/370/ Arofan, Flavio and Hilde presented in the session. Great interest in the topic was shown from participants, of which some potentially could be new and useful contacts for the CDI group.
DDI-CDI will be presented by Arofan and Dan at the forthcoming EDDI Training Fair late november.
2. Model - status - Frozen???
Achim thanked Flavio for his great work on the model. Hilde suggested that the three Signification pattern classes to again become pattern abstractions. This was agreed and fixed by Flavio during the meeting.
3. Organization/timing of field-level doc, diagrams, and updates
A group consisting of Arofan, Flavio, Hilde and Jay will check the latest changes to the model. They will meet on November 5th in order to coordinate.
4. Production flows/tools
Achim will work on the documentation delivery tool this week, so that documentation from the tool can be used to evaluate the model. Production flow/tools topics will be discussed further after the release.
5. RDF binding work?
It was agreed that a sub-group consisting of Achim will focus on the RDF representation work (OWL, JSON-LD).
6. Other looking forward (based on Dagstuhl outcomes)
Minutes Virtual CDI meeting 2021-10-20
Arofan, Dan, Flavio, Hilde, Jay, Wendy
1. Admin (if any)
Arofan informed about a proposal to establish an informal liaison between W3C and the DDI Alliance regarding RDF/JSON-LD. This will be brought up with on the Scientific Board. Arofan suggested Hilde to act as a contact point from the side of the Alliance.
2. Model - status
Flavio has implemented suggestions from Dagstuhl in the model.
He will now look into solutions for requirements from the Unit of Measures group as well as to implement solutions for DMT-273 (Instance Variable’s relationship to a source).
A new version of the model covering the above has been provided post meeting (the frozen one?)
3. Organization/timing of field-level doc, diagrams, and updates
The planning of this will start when a frozen version of the model is available.
4. Looking forward (based on Dagstuhl outcomes)
A set of deliverables is planned to be made available post-release:
RDF syntax representation (using the UML subset), description of use-cases, alignment with other standards etc..
Arofan suggested that the CDI core group work continues to meet every second week, and that small temporary groups are formed, focusing on specific deliverables and bringing in external experts. Model maintenance also needs to be addressed.
Flavio proposed that new ideas, collaborations and standards alignments should be addressed through an open and exploratory process, and that relevance for other DDI products should be brought into the picture.
Wendy suggested post release use cases and deliverables to be discussed with the TC. Arofan is open to discussions between the CDI group and the TC.
Arofan pointed at four goals for the further CDI work.
Exploratory work, get people involved.
New deliverables should be supportive of the product.
Have real use-cases.
Jay expressed support of this, and pointed at model maintenance, use-cases as basis for deliverables, implementation guides, and promoting of the standard as important aspects of focus.
All agreed that activities should be within the frames of the Scientific Plan.