FOSS4G 2022 | State of GeoNetwork

Summary: “State of GeoNetwork” (FOSS4G 2022)

Maintained releases / version roadmap

• GeoNetwork 3.12: regular monthly (or every ~2 months) releases in 2022, largely maintained by the GeoCat team.
• GeoNetwork 4: actively developed alongside v3.
  • GeoNetwork 4.2 (noted as “4.20” in subtitles): released in May 2022 with key search improvements.
• v3.2.6 focus areas:
  • Improving application quality/testing
  • Enhancing the metadata workflow (described as new in v3)
  • Adapting to INSPIRE validator changes (targeting European users)
• Community work also includes:
  • Accessibility improvements (a later talk planned)

Authentication + security + libraries

• Adding support for a new authentication system:
  • Keycloak and OpenID (mentioned as work by contributors)
• Ongoing security updates:
  • Updating dependency/libraries as needed
• Major technical maintenance:
  • Between v3 and v4, major library upgrades (e.g., Spring and Hibernate/equivalent stack pieces moving to newer major versions)

Search architecture changes (big theme: v3 → v4)

• GeoNetwork 4 switches from Lucene-based “homemade” indexing to Elasticsearch
  • Goal: improve catalog indexing and search behavior.
• GeoNetwork 4.2 main improvement: better multilingual search
  • Handles catalogs with one or multiple languages and mixed/partially translated content
  • Improves search results, including nested query behavior
  • Example behavior: showing child records related to a parent record (API/search improvements)
• Performance tuning after migration
  • Elasticsearch was fast, but additional indexing features made it slower
  • Team is working to improve indexing performance, aiming to finalize before 4.2.1 (likely September)

Catalog UX/API improvements and metadata navigation

• Better configuration and map browsing for multilingual catalogs
  • Ability to pick specific maps depending on language
• Improved record navigation for complex metadata structures
  • Relations between records such as series / collections / groups
  • Navigation through service → source datasets, and following replacements across editions/versions
• Improved link organization in metadata records
  • Distinguishes API downloads vs GIS-oriented files (examples mentioned: S3/QGS/legend files)
  • Better categorization of metadata links to reduce confusion when metadata records contain many URLs
• Catalog attribute concept
  • Sometimes described in downloadable PDF, sometimes displayed directly in the record view with browsable attributes

Publishing support for scientific use (citations + DOI)

• Growing demand to cite datasets in scientific publications
• Improvements in citation-related workflows:
  • Citation formatting in multiple ways
    • HTML text (common simple approach)
    • exchange formats like BibTeX (named in subtitles)
  • Collaboration mentioned with projects/organizations (Canadian and French partners like IFR / BRGM)
• DOI registration from within GeoNetwork
  • Ability to register datasets and create DOIs via the GeoNetwork interface

Semantic portals / focused subsets

• GeoNetwork 4 supports creating focused subsets of records (called “semantic portal”)
  • Example focus areas mentioned: biodiversity, marine, water (and configurable for other themes)

Harvesting and metadata creation from external sources

• Harvesting enhancements:
  • Harvest a “GeoNetwork source” from external open data portals/tools
  • Example harvester target: CKAN Open Data ecosystem (interpreted as a CKAN-based tool)
• Subtitles describe defining how to collect information and convert it into GeoNetwork metadata stored in XML

Editing/batch operations and indexing in map/record previews

• Database editing improvements:
  • Search & replace and batch editing capabilities
• Map/record improvements:
  • From record view, preview WMS
  • Ability to index data and enable full-text search in underlying data
  • Example: searching river basin names (e.g., Africa / “Ner” interpreted) and performing aggregations

Migration tasks and schema plugins

• Work mentioned on closing gaps for Elasticsearch migration, including:
  • INSPIRE atom services integration (mentioned with a referenced collaborator)
• Schema plugin maintenance:
  • Schema plugins are still maintained; example named:
    • ISO profile for North America via GeoCat
• Ongoing maintenance of:
  • Dutch profiles for services/data sets
  • Mention of ISO 19115-3D usage (link to a referenced talk/presentation)
  • DCAT-2 described as an experiment / possibly later plan

Community + sustainability challenges (analysis-style portion)

• GeoNetwork described as a mature/older project (started ~2000) with:
  • Technical challenges: keeping old code stable while addressing vulnerable dependencies
  • Difficulty updating “old libraries” and the need for long-term strategies
• Search direction noted as important:
  • Combined search across metadata and actual data
  • Linking metadata to data emphasized as a key future direction
• Community challenges:
  • Hard to find contributors who can dive into core code
  • Need to attract/retain developers
• Project direction: smaller specialized applications
  • Trend toward separate apps for specific tasks instead of only changing the core:
    • Example: Inspire harvester as a separate app
    • Example: a link analysis tool/app to validate/inspect links in metadata/services
    • Example: UI/editor initiatives for simpler editing

UI/editor direction: GeoNetwork UI vs editor complexity

• Mentioned effort: GeoNetwork UI
  • Focus on user experience and alternative UI approaches for different use cases
• Motivation for a new “simplified editing/facet” experience:
  • Some users need only:
    • INSPIRE-compliant metadata,
    • patrimony/search, or
    • limited schemas
  • Goal: provide a simpler interface to edit metadata without overwhelming complexity
• A participative funding initiative is referenced for simplifying editing
• Also referenced: upcoming related presentations
  • Data Hub, accessibility, INSPIRE portal revamp, link analysis

---

Main speakers / sources (as inferred from subtitles)

• Yun (co-presenter; also referenced as organizing a code sprint and speaking on UI/data-related parts)
• Flor (co-presenter; involved in metadata editing/UI)
• Michelle Gabrielle (accessibility-focused contributor; scheduled for an accessibility-related talk)
• Camp to cam / Camp-to-cam team (mentioned as running SonarCloud for code quality)
• Yury Garet / Ros Garcia / other contributors (mentioned in technical contexts; plus referenced organizations/projects)

Primary talk delivered by the GeoNetwork state-of-project presenters: Yun, with GeoCat/GeoNetwork contributors including Flor and collaborators.