Steering committee Minutes 2025-09-04

org-docs/2025-09-04.md

@@ -0,0 +1,104 @@
+# Steering committee 2025-09-04
+
+* Chair: Eli Chadwick (deputy for Abigail Miller)
+* Minuter: Stian Soiland-Reyes
+* Date: 2025-09-04 19:00 UTC
+
+Committee and observers:
+
+- **Attending**: Peter Sefton, Stian Soiland-Reyes, Carole Goble, Rudolf Wittner, Mike Lynch, Eli Chadwick, Marc Portier, Josh Moore
+- **Absent**: Raul Palma, Daniel Garijo, Abigail Miller
+- **Observers**: None
+- **Apologies**: Abigail Miller (travelling)
+- **Quorum**: Reached (8/11)
+
+## Election of chair and secretary
+
+Eli Chadwick was deputy for Abigail Miller.
+
+
+## Approval of previous minutes
+
+Steering committee minutes of [2025-05-22](2025-05-22.md) were approved by consensus ([#18](https://github.com/ResearchObject/governance/pull/18)).
+
+## Review actions from previous steering committee meeting
+
+- Stian to complete the release of RO-Crate 1.2. **Done**   
+  - PID: https://w3id.org/ro/crate/1.2
+  - DOI: https://doi.org/10.5281/zenodo.13751027 
+  - News item: https://galaxyproject.org/news/2025-06-04-ro-crate-1-2-release/ 
+  - Upgrade guide: https://www.researchobject.org/ro-crate/whats-changed-in-1-2
+  - Quick ref: https://www.researchobject.org/ro-crate/quick-reference  
+- Carole to Schedule RO-Crate engagement strategy meeting with Josh, Abigail, Eli. **Pending**
+  - Review by October before next steering committee
+
+
+## Steering committee membership review
+
+No applications to review according to [membership procedure](CHARTER.md#2-steering-committee).
+
+## Choice of discussion topics
+
+It was agreed to discuss _Machine-actionable RO-Crate schemas & validation_, while _Engagement strategy_ discussion was postponed. As a precursor, agreement on how RO-Crate v2 work should be structured is needed.
+
+## Initial planning of features to move forward for RO-Crate v2
+
+The committee notes that [RO-Crate 1.2](https://w3id.org/ro/crate/1.2) constitutes a stable long-term release, and should be used by most implementers.
+
+This allows the RO-Crate steering committee and editorial team to focus on a paceful development of RO-Crate 2.0. Previous steering committee [2025-05-22](2025-05-22.md#vote-on-ro-crate-release) noted:
+
+> The steering committee noted that the expanded 1.2 specification have unfortunately became harder to navigate, and next version 2 needs to focus on editorial cleaning up, **simplifying** for the purpose of implementation and validation. A **modular approach** should be sought, with **feature-full examples** or **benchmark suite** or RO-Crates for programmatically measuring feature completeness.
+
+Brainstorming ideas for RO-Crate v2 have been collected by the community in GitHub under [Milestone RO-Crate 2.0](https://github.com/ResearchObject/ro-crate/milestone/4). The committee reviewed these below.
+
+### Simplicity as core principle
+
+The committee reflected on the need for simplicity in the specification ([ro-crate#386](https://github.com/ResearchObject/ro-crate/issues/386)) and formalised this principle:
+
+**Agreed**: Principle of making RO-Crate 2 smaller and concentrate on formalising the named rules.
+
+### Examples and conformance tests
+
+The need for detailed conformance tests was raised by [[ro-crate#387](https://github.com/ResearchObject/ro-crate/issues/387)], the committee suggest a _“build a test as you go along”_ principle for each of above rules. Josh Moore highlighted the success of such a Request-for-Comment (RFC) like approach [used by the OME-Zarr approach](https://ngff.openmicroscopy.org/rfc/1/), where each proposal is a separate file in GitHub with its own justification, examples and tests.
+
+**Agreed**: RO-Crate 2 rules to be accompanied by positive and negative examples.
+
+The committee agreed on the principle used by other standard organisations of having at least two working implementation of each feature. If a test is not good enough, or implementations disagree in their approach, then that RFC should not be approved.
+
+**Agreed**: Tests of RO-Crate 2 specification to be included by default, with a minimum of two people implementing.
+
+### Entry-points for RO-Crate profiles
+
+The suggestion to have multiple entry-points for an RO-Crate in [ro-crate#371](https://github.com/ResearchObject/ro-crate/issues/371) was discussed. Some RO-Crate profiles can be (re)used with different entities in focus. 
+
+Currently, [RO-Crate 1.2 says profiles](https://www.researchobject.org/ro-crate/specification/1.2/profiles.html) can be applied on the root dataset, with multiple `conformsTo` indicating which profiles.
+Each profile adds their constraints for the RO-Crate as a whole. However, if a profile constrains the root itself, this can prevent profile co-existance in the same crate, in particular if `mainEntity` requirements are conflicting.
+
+**Agreed**: Profile conformance in RO-Crate 2 can be applied at any level of the RO-Crate, not just at root.
+
+The committee notes that a profile may in theory extend the RO-Crate 1.2 specification to be applicable to any entity (e.g. a nested `Dataset`), not just at the root level.
+
+**Agreed**: To add an RO-Crate 1.2 implementation note to highlight how a profile can allow itself to be applied not just at root level.
+
+It was discussed if a single profile could be applicable to multiple types of entities (e.g. either a `DataCatalog` or `Dataset`), but no specific agreement was reached, as this moves into the area of "mix-in" modules for profiles which RO-Crate 2 will have to specify semantics of.
+
+It was _not_ agreed how implementation notes for 1.2 should be published, but it has in previous community meetings been suggested to start a section of implementation notes on the website, independent of RO-Crate specification versions.
+
+## Machine-actionable RO-Crate schemas & validation
+
+Peter Sefton has since 2025-07-10 prepared [Notes for RO-Crate Schemas and Machine Actionable profiles](https://docs.google.com/document/d/17WRkGPIGtoQoSPlTbStBKUyHTzjrOZb620S1gdk0ei8/edit?tab=t.0#heading=h.mueeifp72zyz) (later named _RO-Crate-MAPS_). Some comments were added in the document, but Peter requests further engagement.
+
+A Working group is to be established, and initial discussion is gathered in [ro-crate#442](https://github.com/ResearchObject/ro-crate/issues/442).
+
+
+## Any other business
+
+- Josh Moore highlights US NSF award for [Frameworks-Imaging-PHD](https://www.nsf.gov/awardsearch/show-award/?AWD_ID=2513921): _Empowering data reuse and reproducibility through microscopy-community-defined Persistent Hardware Descriptors_. The use of [PIDinst](https://www.pidinst.org/) with LinkML gap against RO-Crate.
+- The EOSC Life Science Node is forming, with RO-Crate in focus around workflows. Rob Finn (EMBL-EBI) presented RO-Crate as part of the EOSC LS node presentation at the CORDi conference.
+- The [FAIR Digital Object Conference 2026](https://fairdo.org/fdo-conference-2026-call-for-papers/) has deadline of abstracts 26 October 2025, with several RO-Crate related abstracts planned to be submitted.
+
+
+## New actions
+
+1. Regular check-in at RO-Crate community meetings of steering committee progress and some asynchronous nagging by email/github
+2. Add an RO-Crate 1.2 implementation note to highlight how a profile can allow itself to be applied not just at root level