Jon Udell has mentioned the idea of Syndication-Oriented Architecture a couple of times over the the last year of so. One of the things I’ve been trying to spell outabout UMW Blog is how it in many ways is trying to approximate a Syndication-Oriented Architechture using a very hodgepodge collection of plugins and widgets.
What does this mean? Well, for me it means that a university publishing platform shouldn’t only be limited to the sites created within that system (in our case WordPres Multi-User), but rather should be able to incorporate work that students and faculty may be doing on other, externally hosted services that are RSS-enabled—like, for example, Blogger, WordPress.com, Drupal, Flickr, YouTube, etc.
In other words, folks within the campus community should be able to add their feeds to a publishing platform like UMW Blogs and have what they are doing on their own spaces join the general flow of the syndicated data already in UMW Blogs. They should appear in the site wide RSS feed (or even a single class feed); they should show up in the flow of data on the front page; and they should also be searchable in the WPMu archive. Yet, it shouldn’t necessarily be a republishing of their work to another blog space they need to create and maintain, but rather a quick way to drop of their feed so that their work is discoverable by the UMW Blogs community, leaving a trace of their work that will lead people within the community (or a specific class) back to their own space, wherever it is hosted.
We’ve been experimenting with this in a couple of ways recently using Feed WordPress on the Tags blog for UMW Blogs (the tags blog is automatically created when you install Donncha’s Sitewide Feed Tags Page plugin). For example, I can bring in the all the posts for three course being taught by professor Zach Whalen (which he is hosting on his own site in Drupal) by simply adding the feed for each of the courses. The posts are automatically brought into the Tags blog and the permalink directs the user back to the original post in the Drupal course site. The categories can also be imported, comments can be turned off, and you can assign a specific tag for each of the courses if you want to make the work more visible in the tag cloud. Additionally, the posts will show up both on the front page and in the searchable archive.*
So, here’s another example of this in action, Andy Rush has created a pretty bitchin’ New Media blog, and he is hosting it outside of UMW Blogs. Yet, what he blogs about there is of great use and interest to the UWM Blogs community. So, all we have to do is grab his feed, add it to Feed WordPress in the Tags blog, and tag it New Media. After that all his posts will be automatically syndicated into UMW Blogs, and they will also be tagged New Media so that he might be discovered in the tag cloud. Further more, his posts can all be found in the searchable archive, which is just really the search field for the Tags blog.You can see his posts in the UMW Blogs archive here.
Next step is creating something like a self service widget for Feed WordPress so that mebers of a WPMu community can add their feed (or feeds), tag it, categorize it, and then we’re off to the races. The end result is that anyone can publish anywhere as long as it has a valid feed, and their work can still be accessible and searchable by their community as well as the world at large.
Now, with all that said, exploring and discovering work on UMW Blogs is still not that easy, you have to see the flow in real time, or wade through hundreds and hundreds of posts in the sitewide feed. So, my next post on UMW Blogs will be looking at the experimental work Patrick Murray-John is doing with MIT’s Simile Project which he is documenting vernacularly here and more technically here (I’m like a fish out of wter when it comes to the Semantic Web). Andrew Murphy of Metapizza nails it, “Put in an interested use of metadata and web 3.0 and we’re flying …” That is the logic behind this experiment. How might we be able to make the work happening on UMW Blogs easier to search, browse and filter using Semantic Web tools from the Simile Project like Exhibit? It seems the next logical step to try and open up this space further, and try and make the amazing amount of work being done more apparent and finadable by the community at large.
* Much of this is inspired by Bill Fitzgerald’s Drupal setup that he outlines here, and while Drupal can do this much better than WPMu, I still think the ease and feel of the individual spaces provided by WPMu makes it worth the extra work, but that is really just a matter of preference and comfort level in the end.