A bit more than two weeks after Readeck's initial release, a new version is available.
Before we dive into the details of new features, I'd like to wish everyone a Happy New Year and thank the users who happily installed Readeck and shared their enthusiasm and those who provided very helpful feedback, ideas and fixes.
Thank you!
Without more stalling, here are the few new features.
Share a saved link with friends
This is the main new feature. I was reluctant to implement it at first but finally got convinced that we could find a way of tackling the issue.
Readeck now provides a "share" button that lets you generate a public link. You can then share the link (or the QR code) with your friends and family. It expires after one day but you can generate a new link for the same page if one of your friends missed it the first time.
Improved content extraction
Providing an accurate extraction of articles is a thankless job. This release improves the extraction from Wikipedia and Wikinews. While doing this, a bug was found and fixed in the 0.10.6 release.
A too aggressive cleanup of preformatted text was fixed and some images that were previously ignored are now retrieved.
Interface improvements
For better readability, titles are now slightly smaller. Some websites use images as list item bullets and they now render more appropriately.
On Safari, an interface bug was fixed with menu-like elements showing an unexpected marker.
Browser extension
Readeck's browser extension received some updates and interface improvements. Setting the title works now as expected, adding labels matches Readeck's interface and a character set issue was fixed.
Privacy and security
From day one, Readeck removes scripts and other external resources during content extraction. Images are converted and archived with the page's content. On top of that Readeck forces your browser to reject any external resource.
For example, if by accident a page you saved tries to load a script or a font, your browser will prevent them from being loaded. Anything that does not come from Readeck itself is blocked by your browser. There is one exception to this rule though: videos.
When you save a video link (Youtube, Vimeo, TED…), it displays a video player from the content provider; videos are not archived and retrieved remotely each time you press "play".
This release restricts the origin allowed for the playback to the provider's host only. It comes with a bonus; any Youtube link is automatically played from www.youtube-nocookie.com for improved privacy.
Finally, in the same topic, Readeck now denies robots to index and archive its pages, including anything you share publicly.
ARM container image
Starting with Readeck 0.11, an ARM 64 bit container image is available. Please report if you encounter any issue with the new container images.
About the release frequency
I'll do my best to push new releases frequently but I won't commit (yet) to a strict schedule. As long as important fixes or new features are ready, there will be a new release. I favor frequent but small releases.
Happy reading!
Release 0.11.1 - January 10
This releases provides a fix to the extractor. With some websites, inline HTML tags with only spaces, used for punctuation purposes, were considered empty and removed. Now, only empty or white-space only block tags are removed.
Release 0.11.2 - January 13
This release brings significant improvements to the internal HTTP client configuration, enabling HTTP/2, keep-alive and compression. It improves how an article's images are fetched, avoiding failure with some websites.
This version enables fetching "self texts" and galleries from Reddit and improves the extraction from The New Yorker.