The news site as a stream. A live mockup. Some of the ideas that have been baked into this mockup are listed below.
Note: This has been hacked together in little time by journalist with little programming skills. If you'd like to add some functionality to the mockup, pull requests are very welcome.
What you see here is a mockup for a stream-based home page of a news site by @davidbauer. Navigation is replaced by filters that can be combined in any way the user wishes. Topic overview pages and search results are also displayed right there the stream. By default, items are ordered by relevance (a multi-factor-metric that relies on editorial judgement, but not exclusively). One key element is the time slider on top of the stream. It defines the timespan from which stories are displayed in the stream. It is automatically set in a way so that the user sees the most relevant stories since his last visit, but can of course be adjusted to any timespan or simply to "realtime". It solves the problem that all visitors see the same stories, irrespective of when they last checked the news. When a user decides to read an article, the story is expanded in the stream. There are no separate article pages, but of course, every stream element has its own permalink. If a story is visited via the permalink, it is displayed as an expanded item in a topic-stream, with related articles on top and bottom. To allow for generous article layouts, every stream item (aka article) has a fullscreen mode to display articles in a reading-optimised view. Users can manually select topics, sources and authors they are particularly interested in. The personalised feed displays mainly those stories plus the most relevant ones (to burst the filter bubble). Besides the manual selection, machine learning techniques are applied to further refine the personalised feed. Once you have your stream running in the backend, you can deliver it in the frontend whichever way you like, defaulting it to what best fits your audience. You can also offer the whole stream via an API on a freemium model to encourage others to use your stream. Since the stream is in a one-column layout, it can be easily adjusted for all kinds of screen sizes.