Introducing the day Steve Lay reminded delegates of some of the detail of the call under which the projects had been funded. This included a focus on “building and testing software tools, composite applications and or implementing a data format and standards for to defined specification” – in this case QTI. The three funded projects have built directly on the outcomes of previous toolkits and demonstrator activities of the e-framework.

The morning was given over to a demo from the three teams, from Kingston, Cambridge and Southampton Universities respectively, showing how they interoperated by authoring a question in AQuRAte then storing it in Minibix and finally delivering it through ASDEL.

Although the user-interfaces still need a bit of work, the demo did clearly show how using a standards based approach does lead to interoperable systems and that the shorter, more iterative development funding cycle introduced by JISC can actually work.

In the afternoon there were two breakout sessions one dealing with the technical issues around developing and sustaining an open source community, the other looking innovations in assessment. One message that came through from both sessions was the need for more detailed feedback on what approaches and technologies work in the real world. Perhaps some kind of gap analysis between the tool-set we have just now and the needs of the user community combined with more detailed use cases. I think that this approach would certainly help to roadmap future funding calls in the domain as well as helping inform actually practice.

From the techie side of the discussion there was a general feeling of there still being lots of uncertainty about the development of an open source community. How/will/can the 80:20 rule of useful code be reversed? The JISC open source community is still relatively immature and the motivations for be part of it are generally because developers are being paid to be part of it – not because it is the best option. There was a general feeling that more work is needed to help develop, extend and sustain the community and that it is at quite a critical stage in its life-cycle. One suggestion to help with this was the need for a figure head to lead the community – so if you fancy being Mr/Mrs QTI do let us know:-)

More notes from the day are available for the projects’ discussion list.

Packaging and QTI are intrinsically linked – to share and move questions/items they need to be packaged – preferably in an interoperable format:-) However despite recent developments in both the IMS QTI and CP specifications, due to changes in the structure of IMS working groups there have been no public releases of either specifications for well over a year. This is mainly due to the need for at least two working implementations of a specification before public release. In terms of interoperability, general uptake and usabillity this does seem like a perfectly sensible change. But as ever, life is never quite that simple.

IMS Common Cartridge has come along and has turned into something of a flag-bearer for IMS. This has meant that an awful lot of effort from some of the ‘big’ (or perhaps ‘active’ would be more accurate) members of IMS has been concentrated on the development of CC and not pushing implementation of CP1.2 or the latest version of QTI. A decision was taken early in the development of CC to use older, more widely implemented versions of specifications rather than the latest versions. (It should be noted that this looks like changing as more demands are being made on CC which the newer versions of the specs can achieve.)

So, the day was also an opportunity to reflect on what the current state of play is with IMS and other specification bodies, and to discuss with the community what areas they feel are most important for CETIS to be engaging in. Profiling did surface as something that the JISC elearning development community – particularly in the assessment domain – should be developing further.

In terms of specification updates, our host Steve Lay presented a brief history of QTI and future development plans, Adam Cooper (CETIS) gave a round up from the IMS Quarterly meeting held the week before and Wilbert Kraan (CETIS) gave a round up of packaging developments including non IMS initiatives such as OAI-ORE and IEEE RAMLET. On the implementation side of things Ross MacKenzie and Sarah Wood (OU) took us through their experiences of developing common cartridges for the OpenLearn project and Niall Barr (NB Software) gave an overview of integrating QTI and common cartridge. There was also a very stimulating presentation from Linn van der Zanden (SQA) on a pilot project using wikis and blogs as assessment tools.

Presentations/slidecasts ( including as much discussion as was audible) and MP3s are available from the wiki so if you want to get up to speed on what is happening in the wonderful world of specifications – have a listen. There is also an excellent review of the day over on Rowin’s blog.