Thanks to a tip from Bill Fitzgerald, I checked out a copy of the Leech module for Drupal. Despite the rather bad name, it sounds like it is (or eventually will be) perfect for what I need.
It lets users add their own feeds, and can associate said feeds and subsequently aggregated items wit any of the user’s Organic Groups. That takes care of the Class/Cohort/etc… concepts. Users just create or join the appropriate Organic Groups within the Drupal site, and add whatever feeds they want to whatever Organic Groups they want. They could add subfeeds of a blog to different OGs (say myfancyblog.com/tags/bio680/feed to the Biology 680 OG, and myfancyblog.com/tags/poli544 to the Political Science 544 OG…)
It’s not firing on all cylinders at the moment – I think it’s failing to add items to the specified OG, and some other minor things. But it’s going to be waaaaaay simpler to just roll up my sleeves and help out on Leech.module than to write anything from scratch.
Combined with stuff like the Views module, this could be an insanely powerful tool to help pull disparate feeds for users, spread across the internets, into a cohesive community site that behaves organically depending on the whims of the users. Very cool.
{ 1 trackback }
{ 11 comments… read them below or add one }
That’s an awesome looking sketch. I just wanted to drop a note that we’ve posted about a lot of the work we did with a team aggregator based on Drupal, leech, and a handful of other modules here on our blog.
Ian, thanks. It’s still a rough conceptual piece at the moment. I’m debating the benefits of Leech vs. FeedParser, but the basic concepts will be essentially the same. I’ve got a mockup almost working using Leech, and will try the same with FeedParser. I’d built a prototype/mockup months ago using Aggregator2, which worked pretty well except for the ties to Organic Groups.
Actually, it looks like Leech just underwent some pretty serious development, so I’ll be sure to update my copy of the module.
I have also used the MySite module for some of this functionality – viz allowing users to set up their own feeds.
Good points, Scott. The association of students with the SIS system is separate from the aggregation stuff. I see it as possibly a separate custom module that intermediates between Organic Groups, the user database, and whatever SIS is there. Eduglu would have to work without an SIS, so the intermediary would be optional (which is good, since it doesn’t exist at the moment).
I imagine something that talks to a SIS, gets a list of users, groups, etc… and then works some magic unto the Drupal database to represent it properly. Then, someone logs into Edglu/Drupal and it all just magically works. Or, sans SIS, they can create any Organic Groups they want (a teacher could create an OG, and add her students to it. students could create their own groups and add each other, etc…)
D’Arcy, much as the approach you describe will likely satisfy the truly “loosely coupled” crowd, my suspicion (having only just read about it but not actually witnessed what you describe) is that this wouldn’t make it through the doors of many institutions. In my eyes at least, if the mythical eduglu is actually going to make its way inside those doors, it will need to do some of the basic things that mature CMS do and be slightly more tightly coupled, like be able to coordinate users with whatever central registry/service the institution has, and then based on info in the SIS coordinate user accounts being connected automagically with certain groups they need to have access to. Maybe I am wrong. That was at least my understanding of eduglu, is that it was a cross section of RSS registry service with some filters provided by the institutional structures. It does NOT need to limit the student or instructor to just those feeds, and in fact would be woefully recreating some of the short comings of the CMS if it did, but it needs to provide an automated scaffold to at least that stuff. Or, in other language, offer support for formal learning while not being restricted to only formal opportunities. Cheers, Scott
Ok, glad to see I’m on the same page here. I think you’re right in your approach that it has to do both, and if it’s easier to get it going at first just using organic groups to prove the concept that’s cool. As to the other piece (SIS info and then representation within Drupal groups) I expect if we cast around there’s some existing pieces (DrupalEd? Moodle? Any other PHP-based CMSes?) that might have some reusable code or at least examples. To begin, even just being able to import classlists as comma seperated value lists is a good place to start and seems to be the logical place where all of these projects have begun in the past to deal with SIS data. Once that is working then actual ‘integration’ can be worked on as needed.
Hello, D’Arcy,
Glad the leech module is working out — as Dominik pointed out, MySite has some uses here as well —
We’ve also been looking at integration between the Buddylist and ACL modules — between MySite, ACL and Buddylist module, the Leech module (or feedparser), you have a pretty nice combination.
Cheers,
Bill
Just want to note, that for 5.0, we’re looking to build off the feedparser engine and incorporate Leech features into that.
Reason for this is the framework is much more solid and clean and has a better parser.
Still collaborating on this but please keep this one in mind too
Thanks for the module tips… I’m looking for something that will allow our site users to have the sort of Contacts functionality in flickr, et al- so people can “tag” other users as contacts, and be able to see aggregated content from all in that circle… is this something organic groups can handle?
Feedparser looks cool as well…
with the next revision of leech, there will be a URL profiler in the package. it will allow you to track articles back to their origin and cluster them by their URLs.
this is particularily useful when you subscribe to keyword feeds like http://news.google.com/news?hl=en&ned=us&q=drupal&ie=UTF-8&output=rss but want to surf by actual article sources and not by the feed you subscribed to. it will also have some features like pulling screenshots and technorati data for URLs.
alex
Scott, here’s an updated sketch for what I’m thinking of. It’s a bit incoherent, but I swear it makes sense (to me)…