Author: Nathan Willis
I read lots of blogs and planets every day, and do so from multiple computers, so I value the online feed aggregator. It retrieves new items even while I’m offline, and I don’t have to synchronize between multiple machines. Google Reader has filled this niche for a while now, but I’ve kept an eye out for open source equivalents that I could serve from my own domain instead. That is just what GobbleRSS is: a PHP-based feed aggregator that is simple to set up and run on your own.
You can download GobbleRSS as source code from the project’s Web site. To install it, your Web server must have PHP 4 or later (with CURL support) and MySQL. Some optional features require cron access, and the ability to edit an .htaccess file. If you run GobbleRSS on your own server, these extra requirements should not be an issue, but if you are on a shared hosting provider, check what your account supports.
Installing is a simple matter of un-archiving the PHP files into the directory from which you will run GobbleRSS, editing the configuration file, and creating the GobbleRSS database. You won’t even need to create tables in the database, as GobbleRSS will do that on its own when it is accessed.
If you can edit your own .htaccess file, you can add directives to restrict access to GobbleRSS. At present the app is single-user, and you may not want others to be able to access it, so you can add password authentication or IP-based access restrictions, whichever best fits your needs.
When your installation is complete, you can load up the site in your Web browser and begin to add RSS feeds. If you have too many to add manually, or if you just want to migrate your entire feed collection from another reader, you can import subscriptions via an OPML file, as described on GobbleRSS’s installation page.
Features
GobbleRSS works much like Google Reader; a pane on the left side of the page lists your subscriptions and the number of unread items in each, and the right pane displays items from the currently selected feed.
You can navigate using the same keyboard shortcuts provided by Google Reader — and GobbleRSS even goes one better by listing them on-screen at the bottom of the page. You can star items and export individual articles to the ma.gnolia link-sharing site.
If that were all GobbleRSS did, it would be a pretty good replacement for Google Reader. But there’s more.
Guillaume Boudreau created GobbleRSS because Google Reader lacked one important feature: exporting unread items for offline reading. Boudreau’s inspiration was his daily commute; he wanted a simple way to grab stories with his PDA before leaving home so he could read them disconnected along the way. The normal New Articles page displays headlines, but only loads post content for one article at a time — which is just what you can’t do offline with your PDA. To use this feature, you set a maximum number of new articles to export to your PDA in the main GobbleRSS config file, then retrieve them all at once from a special URL.
Another nicety sorely missing in Google Reader is the ability to recognize the same post when it appears in multiple feeds, and eliminate the duplicates. Some desktop feed readers (like Liferea) do this automatically, which proves useful for reading multiple planet-aggregated feeds; Planet GNOME and Planet KDE, for example, aggregate many of the same bloggers as Planet Fedora and Planet SUSE. GobbleRSS indexes posts using the supposedly unique ID element of each feed format — guid in RSS feeds and id in ATOM. As long as the various Planet sites are correctly configured, it will only keep one copy of each post with the same guid.
Limitations
Since GobbleRSS is a one-man operation, its feature set reflects the author’s interests, and that may mean some feed-reading feature you desire are missing. For instance, exporting selected items to ma.gnolia is handy if you use that service, but if you want alternatives like del.icio.us or Scuttle, you are on your own.
More importantly, though, each installation of GobbleRSS is single-user. You can certainly install multiple copies at different URLs and using different databases, but it is not a true multi-user system. And because the current security model is .htaccess-based, it may not be viable for users who want to run GobbleRSS on shared Web hosting.
Likewise, some features (such as downloading feeds to a PDA or other device) require directly altering the GobbleRSS installation, which works fine for a local install, but is not quite as convenient as the basic features provided by the Web front-end.
Also, GobbleRSS’s layout is visually organized, but its look and feel is utilitarian. Boudreau says that the HTML produced is marked up with classes ready to be styled with CSS, but he hasn’t yet had the time to design a new look.
Boudreau has an unofficial roadmap that might include making such changes if users request them and as his time allows. Beyond that, he says, “this project is pretty much something I build new features into when I need them.”
One intriguing idea he mentions is support for APML, the tag-oriented “attention profiling mark-up language.” He adds, “I’m always open to suggestions too, on features on enhancements.”
Conclusion
If I never needed to use a feed reader on any operating system other than Linux, I would be content to use one of the excellent graphical RSS clients that open source programmers have created. But I want to be able to access my favorite feeds anywhere, anytime, so I need a Web-based solution. Proprietary offerings like Google Reader are nice, but have their shortcomings.
The simple ability to weed out duplicate posts makes GobbleRSS a keeper in my book. Perhaps PDA export is the feature you crave. With an open source project like GobbleRSS, new enhancements could be just an email away. While writing this story, I hit several speed bumps while installing GobbleRSS on a shared Web host, and in every case Boudreau responded to my questions and provided a fix — sometimes within minutes. Try getting that level of support from the Google Reader team.
Categories:
- Reviews
- Internet & WWW