4.1 Innovation for education: oss and infrastructure for nz’s  (Page 9/9)

NZOSVLE was not the only OSS project funded. eXe, which I’m sure Wayne MacKintosh will discuss later. NZOSVLE also worked closely with the Open Source Courseware Initiative in NZ team who were undertaking language pack translations for Moodle. In subsequent rounds, TEC funded the OS Learning Object Repository project, the Open Access Repositories in New Zealand, and the Mahara ePortfolio. Eduforge also came about due to the eCDF. So, OSS has been a very significant theme and I’m forever grateful that TEC created this opportunity to establish OSS as such a large part of the landscape here.

It’s inherent with any such fund that some of the dollars get swallowed up in items such as University overheads, ideas that “seemed good at the time” etc. but I’m really happy to say that the overheads for NZOSVLE were kept at a minimum and that we’ve had a really high success rate with getting quality code upstream into standard releases. For this, and making many of these projects the success they are, BIG thanks to Penny Leach, Martin Langhoff and the rest of the programming team for their massive input, much of it in their own time such has been their passion for what we’re doing.

So, a good team was crucial to making it a good investment, having sound project principles, clear goals and vision. These are the things that make for successful projects. There’s one other critical element that made a relatively small investment deliver such a wide ranging impact. Good timing, e.g. Sandy Britain and Oleg Liber’s work on the pedagogy of LMSs, the options and growing maturity of OSS LMSs, the demand for infrastructure in the sector…a worldwide growing interest in OSS for education. Similarly the recent work on Mahara. I think this is good timing, we need options for OSS ePortfolio systems and I believe what we’re trying to do with Mahara will resonate, early days but we’re focused on getting the foundations right.

Thanks Ken for the discussion, and of course for the shared vision and many lunchtime walks we had when you were here in New Zealand and we were setting up NZOSVLE and Eduforge. It has been a very rewarding experience working with OSS in education these past four years and I feel we’re still at the beginning - there’s so much to do!

Cheers Richard Wyles

13. pmasson - march 31st, 2007 at 5:02 pm

Sorry for the late post…

…I hope there is still time to ask a question.

I am very interested in the recent efforts on developing “Moodle Networks.” While at SUNY one of our challenges was to provide integration between the LMS and the 40 or so disparate Student Information Systems (I believe that’s Student Management Systems in NZ?). Campuses where running Datatel,Sungard Banner, PeopleSoft and even home grown systems. The requirement to us was a single interface where any campus could push student and course information to the LMS to create and populate courses, then provide the SIS with course completion, grades, etc. on the way out. Added to this complexity was that each campus had its own unique ID’s for students and faculty and course/section nomenclature. SUNY’s legacy LMS evolved to include its own SIS, causing, for example, students who used the system to not only enroll in their own campus and course, but enroll again within the SUNY system’s LMS–basically double registration. This obviously caused problems with data integrity between the two systems as students added/dropped, enrolled in the wrong course or sections, etc.

This requirement, a common integration interface across SUNY, was considered a must have, yet we could not find an example of any campus or system that had accomplished this in the U.S. Finally we came across a project out of the UK, SUNIWE ( (External Link) ), where cross-campus enrollment was being developed with uPortal. Based on the activity between uPortal and Sakai, we initially thought we may have found a solution. Unfortunately the uPortal/Sakai collaboration proved less than we had hoped.

In addition, SUNY’s technology decisions where moved out from the LMS and technology groups and made by very senior administration. This group was very uncomfortable with any OSS (this will be the topic of my post) and a commercial provider, Angel, was chosen despite both the technical and university system architectural issues. Angel is now expected to provide (build) this single interface for disparate SIS’ or, perhaps this requirement is no longer considered vital.

Can you please provide more information regarding the “Moodle Networks?” How similar are the campuses that will be contributing courses, sharing students, etc. Do they all share an SIS (SMS)–ether a single instance or at least the same application, student ID’s, course/section ID’s.

Thanks, Patrick

14. richardwyles - march 31st, 2007 at 5:55 pm

Hi Patrick, We’ve got a very similar problem here, lots of disparate SIS. When looking at this seemingly mammoth task, I took a KISS approach which effectively sidelines (perhaps ignores!) the issue. Based on the assumption that students in their institutional LMS have already been authenticated via some means (gnarly SIS or otherwise) we built the authentication federation layer to be between LMSs. Each node of Moodle Networks is enabled to allow students form another node in, down to a student or course level. You can set your Moodle to Hub mode as well which would allow any other node to have a trust relationship with it. Reports are transferred to the host Moodle so that if these subsequently transfer back into a SIS that’s accommodated. In essence we’re extending the classic SIS-LMS relationship to being SIS-LMS + trusted friends and thereby abstracting away the problem of SIS interoperability.

The network is conceived so that the student’s access is through their own institutional gateway - their LMS. naturally, this doesn’t solve all the organisational issues such as John Smith wants to take Viticulture 101 from 3rd party provider. These issues can only be solved with cross-credentialing frameworks and all the people issues, but if achieved then Viticulture 101 would be an offering by Institution A (and thereby exist in their SIS) even though it is actually provided by Institution B. We’re trying to develop a distributed network system (with low requirements for governance overhead) rather than a hub and spoke model.

While the technology side had a few challenges, relative to moving the hearts and minds, it’s the easy part. However, by enabling some possibilities I’m sure some interesting configurations will eventuate and many that we didn’t envisage. The inclusion of a pan-institutional learner-driven ePortfolio system (also with federated authentication) adds to the potential of networked learning opportunities.

cheers, Richard