Skip to content

Latest commit

 

History

History
138 lines (83 loc) · 4.21 KB

CHANGELOG.md

File metadata and controls

138 lines (83 loc) · 4.21 KB

0.3.0 (2023-06-09) -- Leaner is better!

💫 Enhancements and new features

  • The Git remote helper git-remote-osf has been reimplemented using the datalad-annex Git remote helper from datalad-next. This has allowed for removing 25% of the code base with no loss of functionality, and actually gaining features (plus various bug fixes). With this release, also "export-mode" dataset deposits can be cloned directly from OSF. Previously, this was just possible for "annex-mode" deposits that suffer from the lack of human-readability.

🐛 Bug Fixes

  • Git remote helper is now fully functional on Windows too. Fixed #112

  • Repeated pushes are now working properly. Fixed #148

  • Force-pushes are now supported. Fixed #162

🛡 Tests

  • The test battery has been ported to pytest.

  • git annex testremote is now also exercised on Windows.

0.2.3 (Mar 2, 2021) -- Cleanup

🏠 Internal

  • Mainly updating the release on PyPi to remove accidentally included remains of the previous osfclient fork.

0.2.2 (Feb 10, 2021) -- No forks

🏠 Internal

  • The internal fork of osfclient has been removed and a dependency to version 0.0.5 (which comes with all necessary features) was added.

0.2.1 (Feb 1, 2021) -- Bugfix

🐛 Bug Fixes

  • A bug that allowed to clone an OSF project from misshaped URLs of type osf:///some/where/underneath (was treated as osf://). This led to an infinite recursion when installing subdatasets.

🏠 Internal

  • An internal function allows to update existing project metadata

  • Updated internal osfclient fork

📝 Documentation

  • Several tweaks to the documentation

🛡 Tests

  • Changes in continuous integration testing with no implications for users

0.2.0 (Jul 17, 2020) -- More DataLad and OSF integration

💫 Enhancements and new features

This release brings a variety improvements that jointly better utilize DataLad and OSF features

  • Add the ability to query a credential store via DataLad, when no credentials are found in environment variables

  • Add osf-credentials command to more conveniently set and reset OSF credentials for use by DataLad

  • create-sibling-osf can now create public projects

  • OSF projects are now of category data by default and another category can be set via create-sibling-osf --category

  • Assign default OSF project tags to location any and specific datasets via OSF search functionality

  • Add the ability to use OSF projects as git-annex exports or actual annex stores

  • Add git-remote-osf Git remote helper to use an OSF project as a regular Git remote, using osf://<projectid> URLs. Performance can be suboptimal when used with datalad push in DataLad versions up to 0.13.0 (repeated, avoidable Git repository uploads). Fixes have been queue for 0.13.1, and 0.14.0.

  • Ability to datalad clone osf://<projectid> to publish and obtain entire datasets via OSF , without the use of a separate service for Git hosting

🪓 Deprecations, removals, API changes

  • Rename create-sibling-osf --sibling to -s/--name for uniformity with other such DataLad commands

  • Rename create-sibling-osf --mode {annexstore,exporttree} to --mode {annex,export} to match git-annex terminology

🛡 Tests

  • Credential-less read-only access to public datasets

🐛 Bug Fixes

  • User/password authentication used user as password and failed

🏠 Internal

  • Dropped dependency on 7z, archive and compression is now implemented via Python standard library functionality

  • Major documentation overhaul to reflect the new features and changed behavior

0.1 (Jun 18, 2020) -- Sprint!

First implementation of a DataLad extension for exchanging data with and via the Open Science Framework (OSF), completed during the OHBM brainhack 2020.

  • A new git-annex special remote implementation git-annex-remote-osf is included that supports using an OSF project as a classic annex, but also supports exporttree=yes

  • A datalad create-sibling-osf command is provided that can programmatically create OSF projects for dataset publication.