Figure 4.1. 

Richard’s OSS series contribution focused around innovation for education and the infrastructure of New Zealand’s education system.

Richard Wyles is a director and co-founder of Flexible Learning Network Ltd, a private company focused on flexible learning solutions for the education, corporate training and public sectors. For the past four years Richard has been leading national eLearning infrastructure projects in New Zealand, underpinned by open source and particularly Moodle. A full-time development team, now numbering around 10 programmers has been working continuously on Moodle and related open source projects since May 2004. Within a short period of time, Moodle is now the most widely used Learning Management System in New Zealand, particularly in the post-secondary vocational educational sector and increasingly within government sector departments.

To coordinate these efforts and provide a place for stakeholders to contribute, Richard co-founded Eduforge. He currently leads several NZ Government funded projects including piloting a national eLearning Network in New Zealand - this work has led to a single-sign-on framework or Moodle Networks, and the NZ Open Educational Resources project.

The saying goes that necessity is the mother of invention. Innovation is somewhat different, it can be incremental improvements, a new way of using something, or the thinking that underpins radical invention. When it comes to innovation there’s two quite distinct drivers. One is the norm in the proprietary software world - that is supplier side innovation. To differentiate a product a supplier will spend on R&D and commercialise and often protect their innovations with patent law. While this model is reasonably efficient in open competitive markets, a significant problem remains in that it largely ignores end-user or demand-side innovation. I say largely because any successful proprietary software vendor, will of course, take demand signals such as customer feedback into account when designing new releases. The problems are that there are time lags, inefficiencies in communication flow and inherent prioritisation of resources that ignores both niche and emergent need (e.g. Does Blackboard have a Maori language pack?). Patents are also designed to limit the diffusion of innovation and thereby protect the competitive advantage that the innovation provides. Problems drive innovation!

Thinking back to 2003 when I first started getting involved in elearning technology, there was a recognised problem in New Zealand’s education system. eLearning was very unevenly spread and quite understandably. New Zealand is reasonably large in geographical terms - a little bit larger than Britain. However, the population is small at 4 million people and we’re geographically isolated - the distance between Wellington and Sydney is not too far off the distance between London and Moscow. It’s a developed Western nation but unusually the economy is largely reliant on agricultural exports. The education sector is well served with 7 universities, 20 institutes of technology and polytechnics, 3 wananga plus many smaller private training companies. Many of the polytechnics are regionally based, serving smaller more rural population centres.

Figure 4.2. Map depicting LMS market in NZ in 2003

In 2003 there wasn’t a lot of eLearning infrastructure. With an initial consortium of 8 institutions, and a modest amount of government funding (given our goals), we started the New Zealand Open Source Virtual Learning Environment (NZOSVLE) project. Our first recognisable problem was that this project was going to be very hard to manage without some suitable tools to help. After looking about, finding nothing at that time that solved the problem and thinking our need can’t be unique, we came up with the idea of Eduforge. Eduforge delivers the same services as does Sourceforge but with some additional collaboration and communication tools such as project based blogging and wikis. We’ve endeavoured to support the needs of both technologists and others in the education community that may be less technically focused. Indeed, there are many projects hosted on Eduforge that have little to do with software. Eduforge is an open access environment - it is not aligned to any institution, it is free to use and has projects from throughout the world. Eduforge could be described as an accidental outcome of the NZOSVLE project. We’ve made some improvements since first launching in February 2004 and we’ll keep evolving the platform. As a trivial aside, Eduforge is now hosted at a data centre in Dallas, Texas to reduce latency for users in many parts of the world.

Figure 4.3. 

In parallel to the work on Eduforge, we needed to start designing the Virtual Learning Environment (VLE). In was vital to establish some core principles to guide our efforts. Firstly, we weren’t going to fall into the “not invented here” trap. A Learning Management System (LMS) was a natural starting point to the VLE and there were numerous open source options in varying states of maturity. We would select the most promising and focus our resources there. We would not fork the code because, with limited resources, a New Zealand fork would only prove to be more expensive to maintain over time. We would be good open source “citizens”. We were constantly thinking, “will this code get upstream?”

So our selection process included not just the qualities of the architecture and code, important though it is. We were also looking for a good community model to apply our time, energy and resources. Though, of course somewhat dated now, this process was documented: Shortlisting of LMS, Evaluation Part II (focused on pedagogical aspects) and Technical Evaluation. The process took a full 5 months with Moodle selected in May 2004.

In hindsight that decision looks relatively easy but at the time there were no clear leaders. Sakai was only just getting underway, ATutor was brand new, Ilias looked interesting as they had made some headway with SCORM compliance but a small user base, and Moodle had a user base of around 350 installations but no enterprise scale installations. Indeed, without some work, Moodle wouldn’t scale to meet our requirements. We weren’t at all concerned about ticking boxes on the features list. We wanted a robust architecture and a responsive open community.

That first year saw a huge amount of effort in improving the scalability and security of Moodle with Moodle 1.5 being what I’d describe as the first truly enterprise ready open source LMS. There were nervous moments launching Moodle at the Open Polytechnic of New Zealand, with its 35,000 learners and we were doing a hard cutover from an in-house system and a gnarly legacy student records system complicated matters. I did the classic project manager’s trick of being far away in Washington DC on launch day, November 1, 2004.

Since then we’ve continued to devote development efforts to Moodle, but now much more into the feature-set and interoperability aspects with other components of the VLE. Our recent efforts have been on developing Moodle Networks coming out as standard in 1.8.Moodle Networks allows a networked framework of multiple Moodles where users can roam across, using comprehensive Single sign-on (SSO) and transparent remote enrollments. Administrators at the originating Moodle install can see logs of remote activity. You can also run your Moodle in “Hub” mode where any Moodle install can connect and users roam across. The Moodle Network code includes an XML-RPC call dispatcher that can expose the whole Moodle API to trusted hosts.

Why did we do this? Again it is to solve a problem. As stated above, many of our institutions are relatively small, serving small remote populations. To ensure broad access to educational opportunities, cross institutional networking of delivery solves student access as well as economies of scale for the institution. The power of the network rests at the node - by that I mean each institution can quite easily configure their Moodle to network specific courses and enroll some students but not others. Institution A may provide say viticulture to Institution B students but not C etc etc. Authentication is managed, as it currently is, via each enrolling institution. The power of this flexible framework will take a bit of time to unfold as it takes some time to establish the non-technical arrangements of such a network.

Concurrently, we’ve been working on a new ePortfolio system. This is a bit of a departure for us because my preference is to build upon existing code-bases than start from scratch. We had been doing some work with Elgg but we got confronted with a design problem in that we couldn’t address the requirements of all the stakeholders in an ePortfolio system with the current architectures available. Mahara (Maori for thought or reflection) deals with this by having an Artifact, Views (templates to group artifacts) and Communities framework. The user can set the permissions on which communities can have access to which views. Still early days on this but we’re very excited by the potential with Mahara. Multiple institutions are using a shared instance at MyPortfolio.ac.nz and that in itself is very rewarding as that level of collaboration would not have been possible only a year or two ago. You can learn more about Mahara by viewing the documents and we will have a demonstrator up soon. Naturally rich interoperability with Moodle is part of the plan and is currently in development.

Another key part of the VLE is a national network of repositories, both for courseware and research output. This is more recent work but we followed the same successful process when selecting the LMS. The technical review pointed to using Fedora for the OAI-PMH national hub and hosted solution while, with some work, Eprints is a good option for ease of deployment at individual institutions. Enhancements we’ve been making include RSS feeds from Fedora, ratings, add comments, nested collections, a DIY configuration tool for Eprints, and a SRW/U service to be adapted for Fedora which will become the basis of the web front end search on the hub and is adaptable for the likes of FEZ and Moodle. I’m probably getting a bit technical here but the idea is to harvest all of NZ’s research output and make it more easily accessible. In parallel we want courseware repositories to be accessible to tutors/teachers/ instructional designers with easy federated search at the course set-up level.

With leads me on to our work on open educational resources but that’s a whole other story…

In summary, what I’m trying to convey with this post is that we’ve been quite busy building what amounts to some significant national infrastructure for NZ’s education system. I like to think that our innovation is end-user / demand driven which is made possible by working with open source technologies. And because it’s open source we can leverage the innovations of others and vice versa.

Our team at Catalyst, the Flexible Learning Network, and consortium partners in the education sector such as the Open Polytechnic are committed to the open source paradigm. It solves a lot of problems for us. When working with open source solutions, the playing field becomes a lot more level as the aggregation of capital is not such of a factor - ideas and capability become the new currency. And for end-users we can deliver innovations and some fit-for-purpose outcomes not otherwise possible. A small but cogent example is that Moodle now has Maori, Tongan and Samoan language packs - important for our native Pacific Island communities. Which proprietary LMS can boast that?