Replacing Google Reader
by Fabian A. Scherschel
For a large number of information collectors and collators on the internet the current challenge is to find a replacement for the Google Reader service. The H's Fabian Scherschel has looked at what functionality made Google Reader popular and what are the current best alternatives to the Reader experience.
When Google announced it would be shutting down Google Reader, the company induced a strong reaction among a very vocal and influential group of internet users. While Google claims that Reader is not one of its most popular services, at least where usage numbers are concerned, it is used by many people on a daily basis and those users tend to be the "power-users" of the internet, people whose job depends on finding and disseminating large amounts of information on a regular basis. In the reactions to Google's decision, both on social networks and in the many articles written on the topic, one theme was repeated over and over again: Google Reader is an important tool that will be missed. Not everyone might use Google Reader, but a large percentage of consumers of RSS feeds do.
The features that made Google Reader popular are the ability of its users to access it anywhere they have a web browser available and the fact that they do not have to install any software to use it. As such, only a hosted service can replace the product easily, although self-hosted software is an alternative if users are prepared to put in the work of installing the software themselves. Despite not having an official API, Google's service also cultivated an impressive ecosystem of mobile applications, which was an essential part of the product's appeal for many people. This is currently the hardest aspect to replicate in the search for a Google Reader replacement.
Its very popularity in the field of, especially web-based, RSS readers means the demise of Google Reader leaves a sizeable void to be filled. Many software developers simply could not see a way to compete with a free product that worked well for its users; despite this, there are alternatives available, both proprietary and open source as well as web-based and stand-alone. There will also probably be a lot of development work on competitors taking place between now and Reader's scheduled death on 1 July.
Guide to reader attributes
OS — open source
P — proprietary
H — hosted service
SH — self-hosted
This article will give an overview of the most popular alternatives to Google Reader that currently exist, starting with hosted services (open source and proprietary) and progressing to self-hosted, open source alternatives, and one rather unique self-hosted proprietary package. Although the focus here is primarily on web-based software, as we are looking to replicate Google Reader's most important characteristics, we have included a roundup of the native open source RSS readers that are available on the Linux desktop, as some users might want to switch to desktop software instead or use it to complement a web-based service.
Open source alternatives are especially interesting as the open nature of their code negates one of the biggest risks with using a proprietary web service such as Google Reader – even if the company or developer decides to discontinue the product, users can always take the code and set up their own instance. This might help users avoid a situation like the one many are faced with now.
NewsBlur (OS, H)
Currently, the closest approximation to Google Reader's feature set, if not design, is probably NewsBlur. Developed by Samuel Clay, the software itself is open source – released under the MIT License – and Clay provides a hosted version for $24 a year. Free accounts are available but are limited to 64 news feeds and less than 10 news items in the reader view at any time. When the news hit that Google Reader would be shutting down, NewsBlur doubled its user base in just a few hours and the developer has been trying to deal with the scaling issues resulting from this influx of users ever since. Even though Clay had to temporarily disable new registrations and had to move his infrastructure to a new hosting provider and a new database backend, the service has now started to stabilise and is usable on a day-to-day basis, even if it is still noticeably slower than Google's service. Clay has detailed the challenges he faced immediately after Google's announcement on the NewsBlur blog.
NewsBlur's interface looks very different from Google Reader's clean, understated lines, but it encompasses largely the same functionality. Users can add feeds to folders, bookmark items and scroll through their timeline smoothly to scan headlines quickly. Features that go beyond what Google Reader offers are the ability to share stories to a public "blurblog", the ability to follow other's blurblogs, and an adaptive filter system that users can use to train the reader in which topics, authors and publications they prefer. This filter system results in an additional reader view that sorts stories not by age or by feed but by the relevance to the user's interests. Android and iOS applications for NewsBlur are also available as open source.
Users who prefer the ease of a service that is hosted for them can pay for NewsBlur with the added security of the knowledge that they can always download the software for themselves and set up their own instance if they so choose. Users who do not want to pay for such a service or who prefer to set the software up on their own servers in the first place can download the code and install it themselves. NewsBlur is based on the Django web application framework, as well as Celery and RabbitMQ on the server side, and uses jQuery and Underscore.js for its front end. The data is stored in a combination of MongoDB and PostgreSQL. More information on how to install and run NewsBlur is available from its README file on GitHub.