Releases: cedricbonhomme/newspipe
Releases · cedricbonhomme/newspipe
Release 9.0
New
- email address of users are no longer stored in the database. Only a
nickname and a password are required to create an account; - deployment is now done with poetry;
- removed a lot bad JavaScript code.
Improvements
- feeds are now retrieved with requests. feedparser is only used for the
parsing; - improvements to the crawler.
Release 8.0
New:
- you can now manage your bookmarks with Newspipe;
- a bookmarklet is available in order to quickly bookmark Web pages;
- import of bookmarks from Pinboard (JSON export);
- new logo.
Improvements:
- refactoring and code cleaning;
- improved Heroku auto-deploy functionality.
Fixes:
- crawler: argument of type 'NoneType' is not iterable in html.unescape().
Release 7.1.3
New
- a template for the articles of public feeds has been added;
- the template of the feeds is now also used for users who are not authenticated (if the feed is not private);
- tags of articles are now displayed in the UI.
Improvements
- it is now possible to configure the feeds refresh interval (in minutes) for the crawler (even if the crawler is used with cron);
- various improvements to the UI of the public profile page;
- simpler format for the logs when the application is running on Heroku.
Release 7.1.2
New
- the new name of JARR is now Newspipe;
- the user can now add its twitter link through the profile page;
- it is now possible to edit the visibility of a feed (if it should be listed in the list of the user's public profile);
- tags of articles are now retrieved in order to use k-means clustering on tags (will be faster than on the article's content);
- various improvements to the crawler (test if an article should be updated and better use of coroutines).
Improvements
- improved the layout of the profile page;
- the React.js page now only lists the feeds with unread articles by default;
- improvements for the crawler.
Removal
- removed the 'refresh_rate' column from the user table;
- removed the export to HTML webzine functionality.
Release 7.1.1
Improvements:
* improved the installation script;
* improved the deployment process with Vagrant.
Fixes:
* fixed a bug that occurred when deleting a user who has categories with feeds in it;
* fixed a bug that occured when the user wants to use SQLite.
Release 7.1
New:
- public profile page for users (private by default);
- popular page: displays the most popular feeds recentlty added by the users;
- new logo.
Improvements:
- the profile edition page has been improved;
- the generation of the tag cloud has been improved (faster).
Fixes:
- fixed a bug when loading the list of stop words.
Release 7.0
New:
- redoing entierly the home page with react;
- implementing category;
- the classic crawler is now taking into account updated articles from feeds;
- support of HTTP proxy has been removed;
- article URL resolving has been removed (wasn't used);
- improvement of the management of users in the dashboard;
- account confirmation is now achieved with a token that expires in a specified time.
Improvements:
- Code re-arangement: move all code to /src/
6.7 (2015-07-21)
New:
- a filter mechanism for feeds has been added (PR #14);
- icon of feeds is now an url retrieved from the feed or the site link (PR #15).
Improvements: - improvements for the bookmarklet (PR #16 and PR #18);
- performance improvements (display of the /feed page);
- enhancements for the two crawlers;
- enhancements of the UI (PR #14);
- misc changes to the models and controllers.
6.6 (2015-06-02)
New:
- it is now possible to sort the list of articles by feed title or
article title; - added base unittests.
Improvements: - fixed some minor bugs;
- improved the asyncio crawler;
- automatically use the good Python executable for the asyncio crawler;
- improved controllers (enforcing better use of user_id);
- the search is now case insensitive.
6.5.5 (2015-04-22)
The full text search powered by Whoosh has been removed.