A report from the first day of this conference at the OU.

Martin Dougiamas’ keynote began as a sort of “State of the Union” address, telling us how widely used it is etc. But mostly he spoke about what’s new in 1.9, and where Moodle will be heading after that.

The major new features in 1.9 are:

  • Gradebook – completely redesigned. Hard to explain briefly what it now offers
  • Outcomes – can create a list of standard outcomes, assign selected ones to activities, and attach them to a scale so they can be ‘graded’
  • Groupings – groups of groups so you can control which groups apply to which activity
  • Performance – problem with slowness in large installations in 1.8, now fixed
  • Tags – way of tagging users (and resources/activities?)

Beyond that, 2.0+ will focus on repository and portfolio aspects. There will be simple methods for bringing in resources from repositories and for exporting them to portfolios. The idea of ‘community hubs’ will develop, allowing sharing of resources between different instances of Moodle on different sites. Also there will be more conditional activities, dependent on outcomes from previous activities.

However, these 2.0+ aspects will wait until after a period of consolidation – where 1.9 will develop through 1.9.x versions to fix all outstanding bugs, streamline and tidy up existing code, so that 2.0 starts from a robust base.

My highlights from the rest of the day:

Nicolas Connault from Moodle HQ spoke about the need to build automated ‘unit testing’ (i.e. testing of individual methods, using ‘mock database objects’ where needed) into code development. So, when you write a new class or method, you write the testing suite to go with it at the same time. Unfortunately it doesn’t work very well for debugging existing code that may not be object-oriented.

Chris Sangwin from Uni Birmingham demonstrated a Moodle version of their STACK assessment system. This allows Maths questions to be created where the student can submit an algebraic answer, which can be assessed according to an algorithm specified by the teacher. For example, you can ask them to enter an ‘even function’ (one that is symmetrical in the dependent axis) and it will be able to determine whether their answer (which could be anything) is indeed even. This is all open source and they want people to test it.

A report on tomorrow to follow…