- Functionality to create a reference list for datasets. This allows users to add a list of DOIs referencing paper, and other datasets that are associated with the dataset.
- Inclusion of support for LSM, JPX, JP2 file formats.
- Provide support for filtering datasets by the collection they belong to.
- Resolving DOIs through DOI.org
- Improvements to file-handling during upload and processing
- Improvements to publishing workflow UX
- Added sorting to records and search results pages
- Enabled Publishers to upload files when dataset is under review for publishing.
- Filter datasets by "Permission". That is, users can filter the dataset list on their permissions to the datasets (i.e. Owner, Manager, Editor, Viewer).
- Email notifications for the embargo publishing process.
- Workflow for requesting early access to embargoed/pre-published datasets on Blackfynn Discover.
- Authentication workflow on Blackfynn Discover to provide access to embargoed datasets
- Scoped autocomplete and suggestions in metadata record search.
- Resolved an edge-case where a deadlock could prevent successful processing of uploaded data.
- Link to most recent published version of a dataset on a page for a removed published dataset.
- Updated copy that is used for Dataset Citations and integration with DataCite.
- Improved responsiveness when querying larger record-sets (> 100,000 records)
- Allow filtering datasets by Permission level. This feature expands the ways in which users can filter the list of datasets that they have access to. Users can now select to display only datasets that they "own", or that they have "manager","edit","read" permissions on.
- If you are manager for a dataset, you can now see all such datasets on the listing page of the Blackfynn platform.
- Datasets will have their "Dataset Date" value updated when associated models and/or records are updated.
- When a dataset is released from embargo, owners and collaborators will receive an email notification.
- DataCite records now reference "Blackfynn Discover" instead of "Blackfynn.
- When searching across datasets with greater than 100,000 records, results will be returned unsorted to improve response times.
- Autocomplete filters will only produce suggestions for models, properties, and values for the current dataset, making search more intuitive.
- Adding a new contributor works as expected now.
- Navigating to a non-existent organization or dataset will gracefully redirect to the Blackfynn login view.
- High priority bug fix in the Discover publish workflow.
- High priority bug fix around file uploading.
- Minor bug fix for dataset banner images disappearing after using the platform for an extended period of time.
- Creating, updating and deleting files, packages, models, records, and properties in a dataset now changes the
updatedAt timestamp on the dataset. Previously this timestamp was only updated when the dataset name, description, etc were changed.
- Optimized listing records for datasets that have millions of records.
- Fixed possibly request ambiguity when a user belongs to multiple organizations.
- Fixed a bug preventing dataset owners from re-requesting the release of an embargoed dataset after the dataset is rejected.
- Datasets, packages, files, models and properties now enforce stricter naming conventions to ensure that file names do not include characters that cause problems with file systems.
- Datasets that have been rejected for publication can now be withdrawn from the publication process entirely.
- Dataset owners can request to release embargoed datasets before the original embargo release date.
- The "Published Datasets" tab in the platform now shows all datasets that are published to Discover from your organization. Previously this only showed datasets that were already shared with you.
- Most API endpoints now return JSON-formatted error responses instead of raw text bodies.
/datasets/paginated endpoint has a new
withRole query parameter which can be used to filter datasets by the permission of the current user.
- Members of the Publisher team were not able to withdraw their own datasets from the publication process.
- Fixed race conditions that could cause large file uploads to intermittently fail on the backed
- Fixed a bug that made it impossible to remove ORCIDs from external collaborators
- Datasets can now be embargoed prior to publication. The files and metadata for an embargoed dataset are published to a private location, and will be made publicly available on Discover on a set date. See https://help.blackfynn.com/en/articles/2239182-submitting-datasets-for-publication-on-blackfynn for more information about this feature.
- Members of the Publisher team can edit datasets which are locked for publication
- Published datasets can now be filtered by organization ID using the
- Searching within metadata array properties now works as expected
- Fixed uploads errors for certain GZIP-encoded files
- Fixed uploads using the Blackfynn CLI Agent for filenames containing a
- Fixed a display bug where external contributors were incorrectly listed as the owner of a dataset
- After renaming packages, downloading files from the package now correctly use the new package name
- Fixed an authorization bug that was causing invalid requests to return with a 500 error.
- Users can now ignore files in a dataset during publication. See full documentation at https://help.blackfynn.com/en/articles/428423
- When creating metadata records for a model, the "title" property must now be set on the record.
- Fixed an issue with package renaming: after deleting a package from the platform, it was not possible to rename another package using the name of the deleted file.
- Fixed a bug when creating contributors linked to Blackfynn users
- In some cases, downloading images from the platform created a ZIP archive containing duplicate files. This has been resolved.
- The legacy tabular viewer has been removed from the Blackfynn app