Community and conformance
Wilbert Kraan, CETIS staff
April 05, 2004

It's the quintessential paradox of technology standards: enforcing a technical standard, while supporting the peculiar demands of a specific community. Such an exercise in cake possession and consumption is now undertaken by the EU Framework Programme 6 (FP6) sponsored TELCERT project, who aim to give national and/or occupational communities the tools to define their own application profiles, and the means to test conformance to them.

The idea behind any interoperability standard is to tighten the structure of information (learning content, resource lists, ePortfolios, whatever) to such an extent that software tools can predict exactly what any specific instance looks like, and behaves like. The tighter, the better. And if you slap a conformance testing regime on top, you close down any loopholes as well.

At the same time, no community uses information in exactly the same way. Some may have no use for, say, bilingual e-learning content, and don't want to have to build support for such a feature into their tools. In other cases, the kinds of qualification categories that go into an ePortfolio's transcript part is necessarily peculiar to the local education system.

Hence the fact that global e-learning specifications such as IMS's have been very general, abstract and almost impossible to conformance test. Witness also the fact that the only successful e-learning specification conformance testing programmes to date have been those that focus on tight application profiles for specific communities: ADL's SCORM, for example.

The trick, then, is to make it easier for communities to specify what exactly they need from an international specification, so that interoperability within the community is more seamless, while not shutting the door to the rest of the world.

The recently started TELCERT project aims to attack that problem from a variety of angles: the Italian CNR/ISTI, for example, will contribute expertise on state-of-the-art techniques for deriving test suites from newish technologies such as UML and XML.

On a more immediately applicable level, the University of Koblenz-Landau will be making tools that can import, create and edit profiles. Using the tool, and the workshops provided by eLoki, communities can switch features of particular specifications on and off, tighten up on others, change vocabularies etc.

Once defined, exising content can be converted to conform to the new application profile using a version of Bolton Institute's RELOAD tool1. RELOAD will be fine for editing new content as well, of course.

Finally, and this is the crunch point, the Open Group will be tweaking its existing general purpose XML conformance test engine technology to both suit the requirements of e-learning material, as well as incorporate new findings that the TELCERT partners discover.

Though the long term business model still needs to be worked out, one can expect that the most likely is that a community designates a conformance committee and/or a certifying authority to define the application profile and the conformance rules; more or less for free. A test authority such as the Open Group will then do the actual testing- for a fee, to be paid by e-learning tool or content vendors. This is not inevitable, of course, as many communities may choose to sponsor the whole process.

One other thing to bear in mind is that 'spec + conformance test' is not the only or always the best way of ensuring interoperability. Interoperability test events that bring together different vendors' tools, for example, can quickly resolve a lot of practical interoperability issues that you wouldn't otherwise find. Still, for ensuring content interoperability, a conformance testing seal-of-approval can be a good way of preventing nasty surprises or expensive content re-engineering.


Though most TELCERT partners are busy setting up webpages, most can be found at the Open Group's TELCERT site at the moment.

1. CETIS is managed by Bolton Institute.