v0.6.1

Hotfix to fix a major issue with search pagination. Also fixes issues with news object displays.

To upgrade from <v0.6.0, follow the instructions from v0.6.0, but pull this tag.

If you have already upgraded to v0.6.0, you can perform the following steps from the ourchive folder, with your virtual environment activated:

git fetch && git checkout v0.6.1
cd ourchive_app/
python manage.py collectstatic
sudo service gunicorn restart

v0.6.0

Bugfixes

• Fixed a number of deploy/setup bugs, including users without upload permissions being unable to comment and new users' default icons 404ing.
• Fixed issues with buggy search logic
• Fixed issues with content page updating
• Fixed UI and functional issues with date created/updated functionality

Features

• Added Language as a foreign key object which can be assigned to works and collections
• Implemented saved searches, so users can persist specific search terms and facets and load searches from the homepage search view
• Added series and anthologies, which are ordered and unordered collections of works
• Implemented chive co-owners and updated logic for approval workflow
• Implemented search groups to give admins control over faceted filter views
• Implemented homepage browsing view, including a new News view
• Added S3 support for uploads

UI

• Began implementing new look & feel, starting with work search tiles and work detail view
• Migrated icons to Tabler
• Updated font

Admin/Project Work

• Migrated repos to OurchiveIO org
• Upgraded technical documentation
• Added an e2e testing project and management command to seed data
• Added a management command for initializing required data such as settings
• Refactored models: all formerly API models are now in ourchive_app.core; this will support full removal of circular dependencies in v0.7.0.
• Migrated project domain to Linode
• Migrated project planning to a self-hosted Taiga instance

Notes/Known Issues

• The project is maturing to the point that we have code debt and known points of inefficiency, such as search. Urgent issues will be hotfixed, but if you notice a particularly bad performance issue or simply have a fix or improvement, we welcome both issues and PRs.

Upgrade Instructions

PLEASE UPGRADE WITH CARE. This release includes a major change to the database schema. If you do these steps out of order your installation will break.

Prerequisites: Ensure you have upgraded to the most recent release prior to this one. The database schema change requires renaming tables prior to running migrations.

As the app user (e.g. ourchive), pull the tag: git fetch && git checkout v0.6.0

Upgrade dependencies. Activate your virtual env and then:

cd ourchive_app
pip install -r requirements.txt

Now you will be executing a management console command to rename tables, running migrations, and restarting the site.

# execute script
python manage.py executesql v0_6_0_core_table_rename

# migrate
python manage.py migrate

# collect static assets
python manage.py collectstatic

# restart app server
sudo service gunicorn restart

Updated Data

Languages can be configured in the admin panel at https://your-ourchive-url.com/admin/core/language

Search groups can be configured at https://your-ourchive-url.com/admin/core/searchgroup

Your homepage will now show a 'news' section. You can populate news at https://your-ourchive-url.com/admin/core/news/. If you check 'embed in homepage', the message will display in the news list as well as on the homepage, below the 'Welcome to Ourchive' message.

Finally, for any tag or attribute types that you want to be browsable from the homepage, navigate to them in the admin panel and check 'Show for browse'.

Thank you

Special thanks to everyone in this Reddit thread, as well as BrightPerigrine and bellisk.

Hotfix: Empty Datasets

This hotfix fixes two issues where attempts to access empty arrays on initial install causes errors.

Maintenance Release: v0.5.5

Important: This release significantly changes search backend code in anticipation of v0.6.0. We recommend upgrading with caution as some regressions may exist.

Changelog

• Updated backend search to standardize facet structure
• Updated backend search exclusion logic
• Migrated frontend search to leverage search objects
• Applied download delete ability to all image/audio uploads in chives
• Added S3 support
• Aligned dropdowns with accessibility standards
• Narrowed chive margins
• Renamed user settings in dropdown
• Updated TinyMCE editor to account for dark mode
• Improved bulk actions for collaborator approval
• Fixed bug with collaborators where removing all collaborators would show blank creator information in chive card
• Updated styling to standardize link behavior
• Fixed bug with draft status toggles not correctly mapping to Django Rest Framework serializer fields

Hotfix: Work & Collection Editing

This hotfix fixes an issue with serializer logic. .update being called on the model via objects.filter rather than the model reference itself being updated caused some data to be overwritten when draft status was saved.

Refer to v0.5.0 for upgrade instructions. This is a single-file change.

v0.5.0

Ourchive Release v0.5.0

Changelog

New

• Work metadata copying to bookmarks - bookmarks now retain the work's title and tags; you can disable this behavior under your profile settings.
• Private bookmarks
• Work subscriptions
• Basic exports for chive data
• Chapter-level video hosting (in addition to work text video embedding)
• Multi-creator works & collections

Updated

• Fixed issues with chive backdating, improved functionality
• Removed importer_requirements.txt from the project (h/t jdm@GitHub for opening an issue on this)
• Link directly to comments on notification page
• Improve admin configuration of image/video/upload permissions
• Improved search, including dynamic range filtering
• Documented Amazon SES email support
• Collections rework
  • Collections have been updated to hold works, rather than bookmarks, and the UI has been updated. This change is intended to support bookmarks' expanded usage as a bookmark-first feature; v0.6 will include the ability to add external URLs as bookmarks, and the ability to add commentary to works you've added to a collection. 'Collections' are intended to allow users to create recommendation lists for works, so a parent-child relationship with bookmarks didn't really make sense. As part of this, the "bookmark detail" view has been removed. Bookmarks are a more private feature; collections are where content curators and reccers can display their analysis.
• OAuth integration beginnings
  • Is OAuth fully functional? No! Functionality is stubbed out and will be live in v0.6.

Fixed

• Preferred download failing on PDF & EPUB
• Videos can't be embedded
• Comments lose functionality on pagination
• Comments don't load correctly when viewing full work on multi-chapter works
• Attributes aren't linked or filterable
• Search behaves unpredictably when filtering on attributes or tags
• Tie token expiration to approval rather than account request date
• Pagination on tag search results defaults to work tab
• Non-rich-text posting broken
• Broken links in documentation (h/t jdm @ GitHub for this fix)

How to upgrade

1. Log onto your server, cd to the site directory, and run: git fetch && git checkout v0.5.0
2. Activate your virtual environment and cd to the project directory (where requirements.txt lives)
3. Update dependencies: pip install -r requirements.txt
4. Double-check your .env file against env.sample
5. Apply migrations: python manage.py migrate
6. Collect static: python manage.py collectstatic
7. Restart your server: systemctl gunicorn restart
8. Restart your scheduler: systemctl apscheduler restart

v04.5.0

This release provides bugfixes and minor improvements as part of working towards version 5.

Bugfixes

• Fixed issue with 'Preferred Download' uploading wherein mimetype was not checked/verified for M4A, M4B, EPUB, PDF, and other allowable formats.

Improvements

• Added backdating to chives (works, bookmarks, collections); post/update dates can now be specified when creating chives.
• Added private bookmarks
• Updated video embed linking so that hotlinked videos e.g. from Tumblr can be embedded in work text
• Refactored TinyMCE into a module to prevent code duplication
• Added pagination to bookmark modal on collections form

Notes

This adds the first migration since MVP's squash of migrations into 0001. If you installed Ourchive previous to the MVP release, you may need to modify migrations on your local setup to get them working. If all else fails, you can remove the API migrations folder and re-add it from this release. Please back up your database prior to performing any migrations.

MVP

This is the MVP release of Ourchive.

See our install guide for detailed installation instructions.