Skip to content

Commit

Permalink
populate and consolidate changelogs. started migration notes, WIP.
Browse files Browse the repository at this point in the history
  • Loading branch information
rpietzsch committed Aug 13, 2024
1 parent 19da6e5 commit 1e34b40
Showing 1 changed file with 172 additions and 15 deletions.
187 changes: 172 additions & 15 deletions docs/release-notes/corporate-memory-24-2/index.md
Original file line number Diff line number Diff line change
Expand Up @@ -16,12 +16,13 @@ Corporate Memory 24.2.0 is the first release in the 24.2 release line.
The highlights of this release are:

- Explore and Author:
-
- All-new, re-written shacl custom UI rendering engine (shacl2) is now generally available and the system default
- Support for Sankey chart type.
- Build:
-
- Quick creation of file based datasets in the workflow editor - dropping files into the workflow editor will automatically create a new dataset.
- Automate:
- Extension to many **import commands** to allow for importing graphs, projects, datasets and vocabularies from the web
- Extension to the **graph validation export** command to produce JUnit XML reports for better integration into CI/CD pipelines
- Extension to many **import commands** to allow for importing graphs, projects, datasets and vocabularies from the web
- Extension to the **graph validation export** command to produce JUnit XML reports for better integration into CI/CD pipelines

This release delivers the following component versions:

Expand All @@ -39,18 +40,163 @@ More detailed information for this release is provided in the next sections.

## eccenca DataIntegration v24.2.0

...
We're excited to bring you the latest update to DataIntegration v24.2, which introduces new features, improvements and bug fixes:

v24.1.0 of DataIntegration adds the following new features:

- Quick creation of file based datasets in the workflow editor.
- Dropping files into the workflow editor will automatically create a new dataset.
- Button to reload all cache activities at once.
- The JDBC dataset can now be configured how it will write multiple values for a single property. A new strategy allows to write multiple rows in this case.
- Python:
- Added `WorkflowContext` that allows plugins to access the workflow identifier as well as the current execution status.
- Added `packageName` attribute to the plugin JSON.

v24.2.0 of DataIntegration introduces the following changes:

- Re-use original parameters of a replaceable dataset if the dataset type matches with the requested one in a variable workflow request.
- Support Turtle files for the graph file upload operator (No chunking supported)
- Transform object header has been separated from its properties to improve visual appearance.
- Rule editors: Add generic path operator to input path tabs.
- When trying to execute an unsaved workflow notify the user that the workflow will be saved with the option to not show the dialog again.
- Re-added reload button to data preview to get updated content on config and data changes.
- Removed URL resource manager. URLs as file names won't be resolved anymore, e.g. for dataset inputs.
- The RDF datasets will always write the schema type for each entity.

v24.2.0 of DataIntegration ships the following fixes:

- Workflow Task: exception message prefixed with wrong name.
- Workflow editor:
- Workflow nodes with only a single dependency output might be executed twice.
- Loose connections from dependency ports not working anymore.
- Allow dependency connection from/to replaceable datasets.
- Drawing dependency connections from input dependency port results in unexpected connections, e.g. connections to data output ports.
- Project page breaks if file resources are missing meta data like size or modified.
- Workflow operator: became invisible after workflow error.
- Always return a 401 (not authorized) instead of a 500 response when refreshing a token has failed with an `invalid_grant` error from Keycloak.
- Rule endpoint does not return new parameters after plugin upgrade.
- Zip file created with macOS Archive Utility containing files with macOS-specific metadata not working in bulk datasets.
- Fixed `overlayEditors` that close even with unsaved changes.
- Consistent navigation behavior or indicate links/buttons that open a new tab/window.
- Replace `highlightedState` properties in workflow editor.
- Rule endpoint does not return new parameters after plugin upgrade.
- Python: Changes in plugin submodule not recognized without restarting DI.
- Dataset API is not fenced against misuse of file parameter.
- Workflow Task: exception message prefixed with wrong name.
- Improve upload icon in DI file upload widget.
- JDBC: H2 driver not found.
- JDBC: Dataset should not write an extra `rdf_type` column.
- JDBC: Cannot rerun Data preview without reopening JDBC dataset update dialog.
- JDBC: naming the URI column `uri` causes `is ambiguous` error.

## eccenca DataManager v24.2.0

...
We are excited to announce the latest update to DataManager v24.2, which introduces new features, improvements and bug fixes.

v24.2.0 of DataManager adds the following new features:

- Support for grouping chart series.
- Support for Sankey chart type.
- A global notifications queue with the dropdown menu besides the user menu.
- Spring Wrapper to Spring Boot 3.2.
- Possibility to add properties that are not visible to shacl2.
- Annotations to shacl2.
- Icons for read-only graphs.
- Validation for the creation items, added links to grid items.
- AC review page.
- Check for simple widget using `defaultResourceViewerIri`.
- "algorithm" param in workspace configuration and `/proxy/:id/resource` queries.
- Notification feature for retrieving the query catalog data when the backend response contains validation errors with the code "DOUBLE_TRIPLE".
- EasyNav
- Create new, directly connected resources via the browse dialogue.
- Improved undo/redo functionality.
- Improved internal structure and robustness.
- Improved selection and searching.
- Adding new connections via shacl shape queries.
- Adding new connections on unsaved nodes.
- Value and UI Query are used in relation manager, if provided.
- Added the multi-source turtle component.

v24.2.0 of DataManager ships the following changes:

- Removed old AC component from "Administration" module, removed feature flag for the new AC module.
- Shacl2 engine is now the system default, removed `shacl2` feature flag, introduced `shacl` feature flag to enforce the use of the old shacl-engine.

v24.2.0 of DataManager ships the following fixes:

- Improved error messages format, now they are more informative and user-friendly with title and details sections
- Workflow triggers reload the page after the workflow is finished.
- E2E tests are fixed, improved and stabilized.
- Resource tags are links.
- Default language for literal editor.
- Layout broken for long labels.
- Align object relation data fetching.
- Labels with more than approx. 24 characters and no white space are now correctly split into two lines.
- Navigation component honors the module settings for `navigationItemsPerPage`.
- Navigation component shows correct pagination for search.
- Unique Visualization names are enforced.
- Fixed UX issues for AC.
- Easynav inverse properties are shown in the correct direction.
- Explore: Turtle tab is visible, even if the user has no write access.

## eccenca DataPlatform v24.2.0

...
We're excited to bring you the latest update to DataPlatform v24.2, which introduces new features, improvements and bug fixes:

v24.2.0 of DataPlatform adds the following new features:

- Added migration of workspace configuration
- workspace/module configurations are stored in JSON content literals in the CMEM Config Graph since v23.1.
- New field <https://vocab.eccenca.com/configuration/json.ver> in workspace configuration graph.
- Endpoint for migrating workspace configurations to current version.
- Extension of actuator info endpoint: Shows current version and number of items to migrate.
- Added SHACL RDF validation view for SHACL batch validations, `POST /api/shacl/validation/batches`
- `validationResultsTargetGraph`: Graph to write rdf validation model into after batch finishes
- `replace`: boolean value on whether to replace the graph (default: false)
- Added option for SHACL Batch run to query target resources with a ignore list for OWL imports, `POST /api/shacl/validation/batches`
- `owlImportsIgnoreList`: A set of graph IRIs which are not queried in the resource selection (i.e. owl imports ignored)
- Added module Access-Control to workspace configuration
- Existing module Administration split into workspace configuration and access control.
- Existing administration module used for workspace configuration (as to avoid migration steps).
- Add alternative endpoint to `/api/shapes/list` called `/api/shapes/listWithValidation`
- return object contains a field for errors in the data which prevent mapping.
- asked behavior results in shape not found responses on shapes which cannot be mapped
- Add `owlImportsResolution` to resource store endpoints
- optional parameter overriding dataplatform setting

v24.2.0 of DataPlatform ships the following changes:

- Update to Apache Jena 5
- All JSON-LD output in JSON-LD 1.1, JSON-LD 1.0 support dropped
- Changed to Access Condition endpoints
- `GET /api/authorization/groups` return IRIs instead of names (including public/admin group)
- `GET /api/authorization splits` pageable parameter into single parameters page, size, sort
- Deactivated graph db change tracking as default
- Extension of ACL review endpoint, response updated with matching access conditions
- Additional SHUI-Property for defining object relation default view
- `shui:viewResourcesWithWidget` with values (`shui:ComplexResourceViewerWidget`, `shui:SimpleResourceViewerWidget`)
- Resource API endpoints `/proxy/{id}/resource` changed
- Additional application parameter proxy.maxCBDStatements (default: 1000000) for limiting amount of statements in memory when loading (S)CBD
- CBD calculation does not include reifications anymore
- Additional one query based algorithm for CBD calculation (algorithm can be selected with optional query parameter algorithm)
- workspace configuration: `apiConfiguration.conciseBoundLoadAlgorithm` (`ITERATIVE`, `QUERY`)
- Endpoint for workspace configuration set `/api/conf/workspaces` falls back on system default
- New default icons for resource, class and properties
- Removal of native stardog integration

v24.2.0 of DataPlatform ships the following fixes:

- Broken workspace configurations fall back to system default workspace
- actuator info endpoint contains field `workspaceConfigurationError` on error
- SHACL controller validation endpoint fixed
- only validates constraints which do not need other data if change-set is given as in memory
- Fixed query rewriting of (named) graphs for users with limited read rights
- From Graphs Rewriting: Prevent rewrite of where clause if no graph variable found

## eccenca Corporate Memory Control (cmemc) v24.2.0

We're excited to bring you the latest update to Corporate Memory Control (cmemc) v24.2, which introduces new features, improvements and bug fixes:

v24.2.0 of Corporate Memory Control (cmemc) adds the following new features:

- `admin store migrate` command
Expand All @@ -75,7 +221,6 @@ v24.2.0 of Corporate Memory Control (cmemc) adds the following new features:
- option `--result-graph` to specifies the graph where the validation results will be written.
- option `--replace` to replace the result graph with new validation results


v24.2.0 of Corporate Memory Control (cmemc) ships the following fixes:

- `graph import` command
Expand All @@ -100,27 +245,39 @@ v24.2.0 of Corporate Memory Control (cmemc) ships the following fixes:

!!! info

TODO

We do not guarantee forward compatibility for configuration, data or projects.
I.e. importing a project created with DataIntegration v24.1.0 into DataIntegration v23.3.0 (or older) might not work.
I.e. importing a project created with DataIntegration v24.2.0 into DataIntegration v24.1.0 (or older) might not work.

Backward compatibility will be ensured or migration paths explained.
I.e. projects created with DataIntegration v23.3.0 can be imported into DataIntegration v24.1.0.
I.e. projects created with DataIntegration v24.1.0 can be imported into DataIntegration v24.2.0.

### DataIntegration

TODO

### DataManager

TODO
- `TODO` feature flag:
- The feature flag `TODO` for the new AC module has been removed as well as the legacy AC component itself. If you used the feature flag `TODO` in your workspace configurations remove it. The old module can not be used any more.
- `shacl2` feature flag:
- Shacl2 engine is now the system default.
- `shacl2` feature flag has been removed. If you used the feature flag `shacl2` in your workspace configurations remove it.
- Use the feature flag `shacl` instead, to enforce the use of the old shacl-engine.

### DataPlatform

TODO
- Due to the update to Apache Jena 5 all JSON-LD output now conforms to JSON-LD 1.1 (was JSON-LD 1.0):
- JSON-LD 1.1 has been designed as a superset of 1.0, so we do not expect any problems, just making you aware that there _might_ be differences when using it with JSON-LD 1.0 and 1.1 processors at the same time (e.g. to compare results)
> JSON-LD 1.1 introduces new features that are compatible with JSON-LD 1.0, but if processed by a JSON-LD 1.0 processor may produce different results (cf. [JSON-LD 1.1 Framing](https://www.w3.org/TR/json-ld11-framing/#features))
- Affected endpoints:
- `POST /proxy/{id}/resource/framed`
- `POST /proxy/{id}/sparql/framed`
- `POST /authorization/conditions/framed`
- `POST+GET /api/queries/jsonld/perform`
- The Access Condition endpoint `GET /api/authorization/groups` returns now IRIs instead of names (including public/admin group).
- Resource API endpoints `/proxy/{id}/resource` changed, the CBD calculation does not include reifications anymore.
- The native stardog integration has been removed. As a stardog user you need to migrate your store configuration to use the generic HTTP-Store configuration option.

### cmemc

No migrations are required going from cmemc 23.1.x to 24.2.x.

0 comments on commit 1e34b40

Please sign in to comment.