Recently I was involved in a project where Moodle had been integrated with Mediawiki for an online community site. The integration of the two systems was limited to a single sign-on facility and consistent visual design through style sheets and graphics. Not full integration, but at first glance a useful start.
But it soon became apparent that this level of integration had some detrimental aspects…
- User permissions in Moodle, which are quite finely-grained, didn’t transfer across to Mediawiki. Not only was it not possible to map roles across the platforms, a user who was given admin rights in Moodle did not automatically become an admin in Mediawiki, even though users seemed to expect this.
- The single sign-off didn’t work as well as single sign-on, so users could log out from one system and still be logged in to the other for some time. Even when the sign-off did work across both, users often perceived they were still logged in as the two systems launched in different windows by default. Users often didn’t fully understand the way web browsers and sessions work and found this confusing.
- Given the lack of real integration between the two systems, the consistent themes caused problems since users didn’t always know which system they were using. Given the limitations of the integration, it would have been better to differentiate the look and feel of the two sites rather than try to make them visually seamless.
Is integration always desirable?
My conclusion was that the project did not really benefit from this limited form of integration, and that the implications for users need to be carefully considered when planning an integration project. In particular, a limited form of integration may be less useful than none at all: the benefits are only achieved when systems are fully integrated, and anything less is likely to cause problems.