The Auricle

In my previous article, Learning Material Repositories: Rafts or Battleships - Part 2, I looked at RSS as one potential approach to aggregating information about, and providing access to, distributed learning resources. But viewing RSS requires a special reader or access to a server-based conversion service … or does it? One of the key features of most Weblog engines is that they automatically generate metadata in RSS or Atom format which makes syndication, and therefore dissemination, about new articles, items, or resources as easy as it gets currently.

But the potential Achille's heel of syndication formats like RSS is the need for a special reader, which is either a separate client application or which is built into a web application. In Auricle, for example, the two drop-down syndication menus are made possible only because the weblog engine we use has an integral RSS interpreter. A normal web browser would simply display the raw RSS, which is not designed for humans to read. Some browser developers are already on the case and so RSS and Atom feeds look like they will be handled 'natively' in future browser versions.

Now for the reasons I explored in Learning Material Repositories: Rafts or Battleships - Part 2, I think low barrier to entry syndication formats like RSS and Atom provide an important balance to the heavy duty battleships of the centralised learning object repositories; so anything which improves access to syndication should be welcome.

For this reason I applaud UKOLN's RSS-xpress RSS to html service and similar initiatives elsewhere, but there is undoubtedly a performance hit for the user (which shows as latency in response) and from UKOLN's perspective I suspect a 'server' hit every time a request to convert an RSS feed to human readable form is received.

So I wanted to explore some alternatives mainly because I wanted to be able to build syndication capability into standard web sites and applications. My basic criteria were that any solution should be responsive, have low latency and be relatively easy to implement.

To cut a long story short I settled on XSLT and CSS as possible solutions and so I've been exploring some of the existing work out there. So before I present my work-in-progress let me first acknowledge a few of those who have trod this path before me.

First up is Sean Burke whose article Making RSS Pretty provides some very useful food for thought. Sean's XSLT didn't work for me because his focus was RSS 2.0 but his article was a good starting point.

Second from the saliva-stimulating UK based Juicy Studio (it's run by Gez Lemon) comes the article XSL Transformations for the RSS Feeds which easily met the short focused article criterion I was looking for. The XSL examples were obviously designed for a specific platform but nevertheless they were clutter free so it was possible to focus down on what they were doing.

Third, we move over the Grack.com where Matthew Mastracci's intriguingly named article RSS + CSS + XSL = Love caught my attention. Again here was a clear articulation of a problem and the provision of a potential solution. Mathew's solution didn't work for me but again it contributed to my knowledge and understanding.

Fourth, is Adal Chiriliuc's tersely titled RSS 1.0 XSLT style sheet. Terse the article title may be but Adal's style sheet worked without a problem.

Now all of these contributed in some way to my thinking but I wanted something minimalist but yet which would produce fairly attractive results. So by way of testing the water I created an XSL stylesheet which transforms the raw Auricle RSS feed (that orange XML icon at the bottom of the right hand menu), which would normally display as a load of markup if displayed in a standard browser. So now, hopefully, if you click on our orange XML icon you may see something slightly more intelligible than normal.

It only needed a very minor adjustment to my weblog engine (information courtesy of Chris Curtis at pMachine) to persuade it to automatically generate the link to the XML stylesheet within my RSS.

What I learned was that XSLT with relatively little effort can do some pretty powerful things like take the machine readable RSS format and repackage it for reading by human beings.

It kind of hurts your brain if you think too much about this but I enter articles into Auricle which are then automatically syndicated into a machine readable format which I then transform back into human readable form using an XSL stylesheet but it's all about improving access and anyway it's the RSS feeds of others I've now got my beady eye on 🙂

If anyone else is interested in undertaking a similar exercise you are welcome to my XSL file which may improve your understanding. If you make any improvements I would be grateful if you could keep me posted in the comments sections of this article.

In the first part of this theme I provided an overview of some of the major developments in learning material repository development and outlined our needs. In this article I try to draw some conclusions and make some decisions about a practical way forward. First let me establish terminology. For our purposes and to avoid a protracted elucidation of 'what is a learning object?', the term as used here includes all digital learning resources, assets, artefacts, and process support tools used to facilitate learning. I'll use the term resources to cover both learning material and learning objects.

Let me be straight up front about this. I believe that rather than focus on the developmental cul-de-sac which is the current mainstream VLEs, the evolving repository models are what really matter; they are the foundations from which other and new models will grow.

Above all we need practical solutions which users will buy into because they perceive them as having a low barrier to entry and use, are efficient, provide benefits over current practices, and above all are flexible (note emphasis).

I'm not suggesting that such learning object repositories can't be complex, modular, scalable, and robust engines behind the scenes, i.e. the battleships of part 1, but that the means by which people and systems make inputs and acquire outputs from such systems shouldn't require at least a first degree in computer or information sciences to exploit the potential.

In previous articles I've suggested that the barrier to entry to e-learning technologies is becoming incredibly high. If you don't believe me then pop over to the CETIS article by Scott Wilson Using the Enterprise SDK to create a minimal IMS Enterprise web service (CETIS 24 June 2004). The argument here is of course that these are enabling infrastructure services from which all else will grow … but boy do IMS specifications appear to build complexity upon complexity!

Whilst recognizing the promise that standards can bring to, what otherwise can be chaos, a large part of this article will be devoted to exploration of a 'standard' and tools which provide a relatively easy way of passing information 'about' learning objects, how to access them and how they are used, rather than technical standards which attempt to rigorously define the structure and nature of a learning object per se. The intention here is not to be a polemic against rigorous technical definitions of learning objects, e.g. IMS Content Packaging, SCORM etc, but instead to recognize that the 'jury is still out' on the nature and benefits (or otherwise) of learning objects particularly within a higher education context. In this article, I'm not challenging the learning object concept but I am taking a pragmatic flexible view of what a learning object is.

Such a flexible view of course means that learning objects from different sources/repositories may have a different 'look and feel' but the benefits of diversity and spread of development effort probably outweigh the insurmountable challenge of achieving global consistency. Greater consistency is of course possible in a single repository or 'walled-garden' or development consortium model.

What this article does challenge, however, is the support for systems which effectively 'lock-in' learning objects, however they are defined, within any single learning object repository. If not we merely repeat what has happened with most learning management systems/virtual learning environments which claim standards compliance but in reality migration remains non-trivial in most cases.

What is required are tools and environments which don't assume that learning objects are necessarily contained within their repository system but which can either 'pull' objects from a variety of sources and repositories, or provide the means by which relevant objects can be found and accessed. What I'm proposing therefore is configurable tools and environments which can find, aggregate, display, and provide access to resources, tools, and services which are potentially distributed over a wide variety of repositories and sources; and which can respond to the user's current learning interests (note emphasis).

The JISC in its own description of the JISC Information Environment (JCIE) has said:
“… the Information Environment as it is proposed here aims to offer the user a more seamless and less complex journey to relevant information and learning resources … It is acknowledged that the Information Environment envisaged for the JISC is ambitious. This is primarily because this has evolved to embrace two key concepts which are by nature semantically and technically complex to advance through a process of investment these are:

• the view that digital resources are inherently distributed and will never be delivered by a single service provider (my emphasis)
• the view that users do not all want to access information in the same way but will require a diverse range of views of resources in order to satisfy their needs. A web based portal or VLE for example may operate as a specific window upon a set of distributed resources.” (my emphasis)

To me, examples of learning objects could be content, activities, or online discussions available from, or mediated via, the web, but also usable on the personal desktop. Because new tools and environments would not need to store content within their own architecture their success would depend on how efficiently and effectively they find, aggregate, or broker both quality-assured and 'non-assured' resources from elsewhere.

If learning object repositories are to have an impact on e-learning the mechanisms by which they communicate with each other and with the end-users needs to be easy to implement. The author has previously highlighted RSS as one easy to implement syndication technology which can be used to enrich a learning environment.

UKOLN's contribution to uptake of RSS is of particular note here by the introduction of their RSSxpress pilot. I'll provide some examples of RSS in actual practice later in the article.

One of the major advocates for syndication technologies in the e-learning arena is undoubtedly Stephen Downes, senior research officer for the Institute for Information Technology, National Research Council, Canada.

I was in the audience at an IMS meeting and conference in Vancouver in February 2003 when Stephen gave his keynote speech No, Really, This is What We Want in which he presented an alternative model to the complex IMS/SCORM content packaging approach. Now credit is due to IMS for providing a platform for an 'alternative' and well argued viewpoint, but I'm not convinced the audience wanted to hear the message which I can distill as “hang on why are you doing all this really complex stuff when we can already do this?”

The 'alternative' model is based on the premise that whilst information about learning objects could be centralised, the actual object itself does not need to be stored in some centralised repository, or even necessarily have a complex descriptive wrapper based on IMS or SCORM specifications. That is not to say that such 'packaging' would not sometimes be desirable; just that it is not always essential.

Many Auricle readers may already be familiar with Downes' prototype DLORN (Distributed Learning Object Repository Network). For those of you who haven't come across it yet there's a useful summary within his notes for a presentation he gave to Open-Education.org Open Education Moving From Concept to Reality (April 2003). To me, Downes merits major plaudits for illustrating how a simple, self-registration, minimum maintenance, online broker system, which utilizes the grassroots syndication technology RSS, has the potential to become a powerful hub of a distributed repository system. DLORN doesn't care about the nature of the underlying systems used to produce the information about learning objects as long as that information is formatted in the relatively simple RSS format which can then be read by the multitude of 'readers' or web applications already out there.

To illustrate the point I've added a new 'Syndicated Learning Objects' menu at the top of Auricle's home page.

Without a DLORN (or equivalent) how would I know what sites were offering their learning objects? Given time and effort, I envisage automatic updates between say a DLORN and an Auricle. I also see some potential for a network of DLORNs based on, say, regions or subject specialisms? A more sophisticated DLORN could even go on to offer resource links and information specific to a particular part of a course or module.

The example below shows three RSS feeds from different sources displayed in one web page. Ok, I've used a frameset to do this, but I'm only wanting to illustrate a point. Because RSS is an XML-based format, feeds cannot be displayed directly in a web page; but here UKOLN’s RSSxpress service has been used to convert the output from an RSS url, e.g. http://www.ch.ic.ac.uk/index.rss into output which can be rendered by a standard web browser.

But all of this may not be music to the ears of those convinced the World needs more sophisticated solutions.

In RSS: Disruptive Technology Hiding in Plain Sight (Seidl FA, 2003) RSS is described as a 'disruptive technology' which will become an attractive alternative to existing solutions.

“Generally, disruptive innovations were technologically straightforward, consisting of off-the-shelf components put together in a product architecture that was often simpler than prior approaches. They offered less of what customers in established markets wanted and so could rarely be initially employed there. They offered a different package of attributes valued only in emerging markets remote from, and unimportant to, the mainstream.” (Christensen C, 2003, The Innovator’s Dilemma, HarperBusiness)

In a more recent polemic by Downes Whither the Semantic Web (Downes S, 17 June 2004) he contends that:

“The most successful part of the Semantic Web is RSS (Rich Site Summary, or Really Simple Syndication). That RSS is successful is beyond dispute; there are millions of sites using RSS worldwide and millions of people reading RSS. But what is ironic is that RSS developed outside the Semantic Web development infrastructure; if the Semantic Web is what the W3C is building (and only that), then RSS is not a part of the Semantic Web.”

Or in RSS: Grassroots Support Leads to Mass Appeal Downes suggests that the bottom-up origins of RSS means that already it is already having a practical effect, although this does not necessarily map to what was intended by infrastructuralists :

“Although RSS is not the semantic Web originally dreamed of in the laboratory, with finely grained and standardized element descriptions and canonical vocabularies, it is a technology that has proved itself, and evolved roughshod, though the much grittier practice of grassroots development. There is, I think, a lesson in that.”

But more optimistically he also suggests that:

“It is only a matter of time before the RSS and OAI worlds merge.”

Harking back to my first article on this theme, the Open Archives Initiative has been the launch point for initiatives such as DSpace.

RSS adoption continues to increase and is now a core part of many services and markets, e.g. weblogs, news syndication, and corporate portals. Weblogs, portals, and RSS are viewed in some quarters as potential parts of a knowledge management infrastructure, e.g. Myst Technology.

“Portal and KM vendors could learn a few tricks from emerging technology segments like RSS, RDF, and the blogging community. These initiatives have stumbled upon [what I consider] the single most important aspect of network dynamics - the discrete addressability of information objects.”

RSS support is already being integrated directly into some e-learning tools/environments. For example CourseForum is web-based e-learning software that enables students to interact by creating, posting, share or discuss course content. The product now supports RSS feeds natively, i.e. without the intervention of other services like UKOLN’s RSSxpress, so that web forums can be enhanced by external information or resources. Other mainstream VLEs tend to offer RSS only at the enterprise or portal level of their products range and then view it only as a news distribution solution. Pity! They are missing a feature which deserves to be more core than peripheral.

There are other examples of RSS being used as a means of learning object dissemination. For example, as shown below the Macricopa Learning eXchange or MLX.

At the time of writing the MLX offers over 900 learning items to its participating colleges with searches being offered as RSS feeds which can integrated into academics' blogs or web sites. At the time of writing, for instance, a search on the MLX for items relating to 'heart' provides an RSS formatted search result of several items one of which is Ventilation and Perfusion on the Web. Selecting this item takes us to a descriptive packaging slip which provides the provenance, original pedagogical purpose/context of the item and a web link to the actual item.

Interestingly, the MLX packaging slip also enables 'sharebacks' (also known as trackbacks) on the item. Trackbacks are one solution to learning about the reuse of learning objects since they provide either comment upon, brief descriptions of who is reusing, or how a learning object is being reused. Further consideration of using weblog trackbacks to provide context for learning objects can be found at either D'Arcy Norman's Learning Commons blog or Alan Levine’s article Backtrack to Trackback.

As shown below, the MLX site offers an example of how one academic integrates MLX's and other syndicated learning objects into his personal humanities-oriented web site.

Important point, the MLX is going open source and so will be available to us all in the near future. It's based on a LAMP model (Linux, Apache, MSQL, PHP) so the cost of getting our hands dirty shouldn't be too high. I don't know about the underlying robustness or quality of code e.t.c. but what has always impressed me is the clarity of concept, its apparent ease of use, and the community oriented spirit of its creators.

I think the MLX solution deserves a lot more attention on this side of the pond than it has been getting. It's not unusual to talk to UK repository oriented folks who have never heard of the MLX. Hopefully this article will help disseminate their good work a little more.

Meanwhile back in the UK, MedWeb at the University of Birmingham provides us with some further examples of RSS disseminated learning objects in use. One example shows how RSS formatted resource discovery can facilitate integration into a host environment (below).

In the syndicated learning object menu at the top of the Auricle home page I link to a second MedWeb example, an RSS feed of prototype MedWeb learning objects; these can be viewed directly in RSS enabled sites like Auricle, a standalone RSS Reader, or converted to html for display in a standard web browser by a conversion service, e.g. UKOLN’s RSS-xpress-Lite.

David Davies of MedWeb is leading the Learning and Teaching Support Network mini-project investigating pragmatic approaches to resource discovery and collaborative development.

RSS feeds, like MedWeb's, can be incorporated into many different types of container, including mainstream virtual learning environments. While I believe this to be an enrichment of the otherwise closed world of the VLE, it does raise the spectre (for vendors) of such VLEs becoming devoid of their own learning objects and merely being relegated to the role of relatively expensive delivery vehicles or front-ends. In such a scenario it becomes a moot point whether it is cost-effective to support a proprietary VLE for this purpose at all:)

It’s also interesting how VLE vendors allocate RSS a relatively minor role. Whilst RSS is enabled via the News module in the Blackboard Portal System, the same does not appear to be the case in the Blackboard Learning System. In the Blackboard Portal a University can syndicate content through the portal (once the user activates the News module on their customizable portal page and selects which of the defined feeds and number of headlines to subscribe to. This solution doesn't appear to offer a native mechanism that allows a user to subscribe to any feed they want, but that's something that Blackboard's Building Blocks technology model may make possible.

I know I repeat myself but what is key in all of the above examples is the recognition that learning objects do not need to be centralised or locked in a single repository to be accessible and useful. Many of the examples do require a flexible interpretation of what a learning object is. RSS is, however, pedagogically neutral and is simply the vehicle which carries information about, and provides access to, the object via a link.

The UK lags behind North America in the development of RSS capable learning object repositories (which could be viewed as an opportunity). Apart from the previously mentioned Distributed Learning Object Repository Network (DLORN) current RSS capable learning object repositories and sources include:

CAREO Campus Alberta Repository of Educational Objects) is a collaboration between the Universities of Alberta and Calgary in co-operation with CANARIE (Canadian Network for the Advancement of Research in Industry and Education), and BELLE (Broadband Enabled Lifelong Learning Environment) …

and …

MERLOT has a long track record as an online learning materials repository now also offers RSS feeds. For example any of the following RSS feeds will return information about the ten most recent MERLOT assets in a particular domain:

• http://rss.merlot.org/publish/Physics.xml
• http://rss.merlot.org/publish/History.xml
• http://rss.merlot.org/publish/Chemistry.xml
• http://rss.merlot.org/publish/Biology.xml

For a summary of other known repositories Scott Leslie’s EdTech Post article is a useful source or Bill Brandon’s blog. Again, a liberal interpretation of what constitutes a learning object is necessary when viewing these examples.

It's even possible to have the results of Amazon searches returned as RSS feeds for incorporation into whatever container is preferred. A similar result can be achieved with Google by using using a third-party, currently free, service Google.RSS but the caveat with any service but particularly free ones is 'here today gone tomorrow' so building course dependencies using such services is probably not wise.

The pity is that some sections of the e-learning community haven't seemed to grasp the potential. For example, the National Learning Network learning materials repository for the post-16 sector appears to eschew RSS feeds as does the UK’s National Electronic Library for Health which apparently only offers its headlines, what's new and 'document of the week' as RSS; not RSS formatted search results (which could be incorporated into an online learning resource).

There's also a new syndication contender on the block called Atom which offers richer features than is 'officially' possible in RSS (with readers becoming available for both) but in this article I'm arguing for a principle rather than a specific solution. Syndication solutions like RSS and Atom meet the 'low barrier to entry' criteria I identified in part 1 and can be used by a variety of containers or clients. They are certainly don't require the design and development effort required for battleships but yet like rafts they can be built and used by the grassroots developer/author while battleships can also release rafts … I hope the metaphor is holding up … if it doesn't make sense read part 1:)

In conclusion, if learning object repositories are to have an impact on e-learning then I've proposed that the mechanisms by which they communicate with each other and with the end-users needs to be easy to implement. I've attempted to highlight RSS as one easy to implement syndication technology which can be used to enrich and populate a learning environment.

An excellent resource with useful links to RSS tools and validators is RSS - A Primer for Publishers and Content Developers by M.Moffat, at EEVL.

A graphical representation of the growth of RSS is usually available from Sydic8.com.

On Tuesday (20 July 2004) I gave the keynote address at the University of Leicester's first e-learning conference. Despite it not being the optimum time for such a gathering the conference was fully subscribed. In this article I highlight some of the work of other presenters as well as provide an overview and links to supporting resources for my own presentation. It was refreshing to see so many academics from across the region and Leicester, a university with a respected research and teaching profile, who were obviously keen to find out what, if anything, e-learning has to offer them.

As well as offering traditional full-time campus-based learning programmes, Leicester is involved in significant flexible and distance learning; I suspect this provides an impetus for pedagogical and logistical diversity that institutions with a traditional campus-based focus may lack, currently. Leicester can only benefit from this. E-Learning is obviously perceived as being of increasing importance to the work of the university as evidenced by the recent appointment of Professor Gilly Salmon (currently UK Open University) as the University of Leicester's Professor of E-Learning and Learning Technologies. Gilly is well known for her work on E-Moderation and activity-based learning, i.e. E-tivities. The appointment of a Professor of E-Learning and Learning Technologies is making a strategic statement; other universtities should perhaps take note.

Two things impressed me immediately.

First, Professor John Fothergill who is both Professor of Engineering and Pro-Vice-Chancellor at the University of Leicester opened the conference. Now John is not just a senior university administrator he is also an actual 'getting his hands dirty' e-learning exponent and delivers his own module on Optical Fibre Communication Engineering via sets of online learning objects (Impatica encoded Powerpoints), supporting materials (books, videos, formative assessments) and tutor-supported discussion boards. Respect! Leading by example is good leadership. So John's 'street cred' was immediately high and his welcome was in part based on his own grounded experiences.

Second, the speaker who followed me was Mark Russel winner of this year's 'e-Tutor' of the year award. Mark is Principal Lecturer in Aerospace, Automotive and Design Engineering at the University of Hertfordshire in the UK. The following extract from his own institution's newsletter summarizes pretty well what he did.

“His submission had been based on a first-year fluid mechanics and thermodynamics module for which he improved the pass rate by 27%. Mark attributes the improvement to a range of electronic technologies he developed to, as he puts it, “…nudge, nurture, challenge and assess students”… The four competition judges agreed that Mark had made innovative use of the University's own 'StudyNet' managed learning environment, Excel (which read the work submitted, marked it and sent back an individualised e-mail based on performance) and such common software features as mail-merge. These had been combined to encourage collaborative and independent learning, to develop the student's confidence, to challenge and extend their learning, and very importantly, to provide feedback on their weekly computer-assisted assessment tasks.”

At the conference Mark described how he issued weekly questions to around 150 students. Although the question stems were all the same the values in the questions were all unique and so, at a stroke, collusion was out. Students submitted their responses to an online aggregator which also encouraged them to submit a hint or tip about the module. These hints/tips were then made available to the whole cohort and those who submitted were rewarded for their contribution by two additional marks. Mark is utilizing assessment as a driver of learning and has automated large parts of the system using a combination of Excel, Outlook and custom scripting to handle the workflow of submission, acknowledgements and fairly terse, but obviously effective, feedback. Like most highly motivated and capable people, however, Mark believes his solution is 'easy' and could be applied to other contexts with little effort. I think what he has done is wonderful but I look forward to examples of how easily it can be introduced elsewhere. Nevertheless, Mark is your man if you want a really grounded example of assessment in practice.

Now on to my contribution to the proceedings.

I chose a suitably cosmic theme 'Terraforming the environment for sustainable e-learning life!' Why? Because the University of Leicester hosts the Space Research Centre which in turn is also a key component of the UK's National Space Centre.

Terraforming was a good launch point for a consideration of how environments sometimes require modification in order to create the conditions for sustainable life. We weren't talking innovation here but change. I also considered briefly the role of failure in contributing to ultimate success, e.g. the UKeU.

In essence I was arguing no more than others, more august than myself, have already argued, i.e. in reality HE is still dominated by a knowledge (or in many cases information) transfer model and so change is frustrated by a desire to replicate this model using information and communication technologies. The result is we amplify many of the current problems instead of finding and exploiting new opportunities and solutions.

In the 'How Not to Terraform' section of my address I argued that we need to realise that e-learning isn't really about technology at all; instead it's a people thing, a communication thing, a process thing. I also indulged myself with a mini polemic on the adoption of 'student-centredness' and 'independent learning' by institutions as a 'flag of convenience' but which is not evidenced by the provision of a support and resource infrastructure to back it up. I also proposed that perhaps we are better at lots of innovation but less good at achieving sustainable change, e.g. what happens to these exciting and innovative projects at the end of their lifetimes or shortly afterwards?

I then went on to propose that success in e-learning (and learning generally) could be guaranteed … but only when there was explicit alignment of learning outcomes, assessment, activities, events and resources. From here I went on to explore the attributes of an effective learning environment and the importance of refocusing to a less content dominated and more active approach to learning. That is not to say that content and resources don't have a major part to play but it is only a part. A brief consideration of the ingredients of a student centred virtual learning environment was followed by some examples of active learning from within the UK and international HE community.

I finished up with a consideration of 'What looks promising on the e-Terraforming front' with a particular emphasis on the wealth of information, case studies, islands of good practice, and focused development funding calls that now exist. A brief mention of the Program in Course Redesign (aka the Pew Learning and Technology Program), Learning Design, the JISC E-Learning Programme, Maricopa Learning eXchange, and JORUM and I was done.

My slides for the address are available in two options (both colour PDF). You can view them as two slides per page or six slides per page. I suggest you right click your mouse to download them to your local computer.

Let me fling away my muddled thinking and 'get with the programme'. No more resources 'locked-in' to e-learning environments, tools, or multimedia applications. I want my educational applications populated from a 'proper' learning object repository … don't I? That was the mission, but here's the findings and thinking so far. We've been involved in a number of e-learning/multimedia projects over the years whose design invariably included resources stored in some type of virtual 'library', archive, or repository. The repository was invariably hard-wired to the application and its virtual status arose simply from the fact it was delivered via a computer.

Now we know better and we really don't want to embark on 'locking in' learning material/resources/objects within monolithic, custom, or proprietary architectures. We want to embrace the world of specifications, standards, open source, and learning object repositories. All well and good but the problem is that the need is now but we're finding that developments are mostly works-in-progress and 'jam tomorrow'.

So let's start off by articulating the need.

We want the material which is utilized by a learning tool, environment or application to be separate from the container(s) in which it can be presented. We want easy authoring, updating, archiving, retrieval and overall management of such material. We also want it to be finding and retrievable by software clients and even software agents/robots as well as humans. Why the latter?

Basically, we don't want to be trapped by the application interface provided by any one vendor and can therefore integrate the material, or at least links to it, within whatever software clients and containers we care to use.

So step forward the brave world of learning object repositories.

We've found some interesting work for example the JISC JORUM project which describes itself as:

“… a repository service for all Further and Higher Education Institutions in the United Kingdom (UK), providing access to materials and encouraging the sharing, re-use and re-purposing of them between teaching staff.”

Interesting although JORUM is, the underlying repository engine appears to be Intrallect's IntraLibrary learning object management system, which, in its public form at least, is an application (although it's architecture makes services feasible). IntraLibrary is described in Digital Repositories: e-Learning for Everyone (PDF file) (Duncan C, 2003) as a “commercial successor” to SESDL the Scottish Staff Development Library.

Also, the JORUM team are well grounded in the political realities of the heterogenous UK Higher Education sector meaning that JORUM will be interacting in an environment in which institutional learning object repositories will also exist. It's unclear currently how JORUM will communicate with such local, regional, and national repositories; and indeed repositories already on the international stage such as MERLOT.

So let's take a look at what digital library systems (as opposed to learning object repositories) such as DSpace and EPrints have to offer.

The DSpace site offers the following description:

“… a groundbreaking digital library system to capture, store, index, preserve, and redistribute the intellectual output of a university’s research faculty in digital formats.”

EPrints offers the more muted:

“… dedicated to opening access to the refereed research literature online through author/institution self-archiving”.

The difficulty for us is that these two initiatives are what they say on the tin. We find it kind of difficult, currently, to visualize how they could be massaged into delivering what we want of a smaller grained, learning materials/resources focused service which is not dependent on interaction with a fixed digital library application interface.

So let's look at what's coming down the slipway. What about the cross-platform LionShare project? LionShare will:

“facilitate legitimate file-sharing among institutions around the world through the use of authenticated Peer-to-Peer (P2P) networking. LionShare technology (which is currently under development) will provide tools for the exchange of academic, personal and work-related materials on an officially sanctioned and secure P2P network among participating groups and institutions around the world.”

LionShare is certainly interesting with betas due for release in Fall 2004 and release candidates in Fall 2005. I've included this project in Auricle's syndicated sites menu.

More peer-to-peer options. After reading the CETIS article Splashing in Ponds and Pools I hot-footed over to the EduSplash site.

The EduSplash site confuses me. On the one hand it describes itself as a:

“… consortium of several educational, private and public, sector organizations to develop an infrastructure for learning object repositories.”

But pop into the news section and we find the last entry was apparently 30 June 2002!

I thought I had entered a 'Marie Celeste' site but perhaps not! … in the download section the latest build of the Splash Java client 🙁 is 28 June 2004?. Now I like the idea of Splash empowering the individual whilst contributing to a community via peer to peer networks of Splash personal repositories (a pond) and to a more general resource network (a pool). But could I find any information about the pond and pool parts of the architecture … no I couldn't. So either its a secret or I'm looking at the wrong site.

I did try out the Splash client and found it a bit slow and flaky but it has some promise. Conceptually, it's seem to me a bit like the JISC Reload tool but one which is aware of the learning objects created by others and which can make available to a community the learning objects it describes … a potentially powerful approach to resource description and sharing … are you listening Reload:)

No learning object repository excursion, no matter how brief, can ignore Fedora. By the way the url is http://www.fedora.info and don't forget that .info at the end of the url or you'll be treated to a totally unrelated assault on your senses.

It seems to me that Fedora is where the big boys and girls go to play.

“The Fedora project was funded by the Andrew W. Mellon Foundation to build an open-source digital object repository management system based on the Flexible Extensible Digital Object and Repository Architecture (Fedora). The new system demonstrates how distributed digital library architecture can be deployed using web-based technologies, including XML and Web services.”

Now that last sentence is actually critical … this system is apparently being built on a service model and is intended to be:

“a general-purpose digital object repository system that can be used in whole or part to support a variety of use cases including: institutional repositories, digital libraries, content management, digital asset management, scholarly publishing, and digital preservation.”

So it is going beyond the DSpace and EPrints model. Fedora is breaking the mould in another direction in its adoption of the METS XML schema, i.e. the Metadata Encoding and Transmission Standard (just when you thought you'd grasped IMS, SCORM et al 🙂

Fedora is undoubtedly powerful enough to keep sofware architects and engineers happy for years, and infrastructuralists will undoubtedly love it. Perhaps its service-oriented architecture is what matters, however, since most normal mortals will only need to be concerned with how they can get the data and information they need from a Fedora engine and leave the high priests and priestesses to administer to the needs of the underlying system. At the time of writing IMS hasn't influenced the design of Fedora but it's likely that both initiatives will need to talk to each other … urgently!

This seems a good point to bring part 1 of this article to a conclusion but first where did the title 'Learning Material Repositories - Rafts or Battleships?' come from?

Well here's an extract from a recent Stephen Downes recent polemic

“In the world of e-learning, meanwhile, the systems and protocols look more and more like jibberish each passing day as every possible requirement from every possible system - whether it makes sense or not - is piled into that tangle of 24-character variable names called Java (none of which will work at all unless you have exactly the right configuration … But the last time I looked people weren't using learning objects in any great number, either in the classroom or (even more so) to support home learning. Gosh, make sure you can float before building a battleship.” (OLDaily 12 July 2004)

What part 2 will explore is how we can float.