In practice, the difficulties faced by a teacher who wishes to adapt an OER may be insurmountable, especially when we consider that even specialists in software development acknowledge modifying software to meet changing requirements to be one of the most challenging problems in software engineering. A further complication is that educational resources can be built using a wide variety of software resources, packages, programming paradigms etc and targetted at different modes of use (mobile, offline, online, collaborative) and platforms (workstations, laptops, tablets etc). Ways of exploiting OERs that seem conceptually straightforward to a teacher may in fact be prohibitively costly if not infeasible on account of the well-known problems of conceptual incompatibity and interoperability. One consequence of this is that, despite the aspiration to exploit and repurpose rich resources that have already been developed, OERs tend to work only in isolation from each other, and new developments almost invariably begin afresh.
This short presentation argues that resolving this strategic issue involves more than political and managerial initiatives. The way in which software is developed is fundamentally ill-suited to meeting the demands for open educational resources that can be realistically deployed and exploited to their fullest advantage.
The CONSTRUIT project  is an ongoing study that aims to address this issue by promoting a new paradigm for developing educational resources. The central focus in CONSTRUIT is on ‘making construals’ (a concept introduced by Gooding ) – the development of live interactive resources that serve as shareable working models. The development of construals is an alternative approach to software development that puts its emphasis upon drawing on expert knowledge of the target educational domain to construct fluid interactive environments in which the agency of teachers and learners can be freely enacted. Such an approach, which is radically different from traditional programming, leads to products that are much more flexible in character. Examples of construals include sorting algorithms, electrical circuits and models of the solar system.
For instance, a single construal can be the basis for myriad educational applications that can be derived from expert knowledge of the educational domain rather than professional programming expertise, one construal can be readily remixed with other construals without wholesale reconfiguration and modification by a software specialist, and construals can be developed and transformed into educational resources through live, informal and empirical collaboration, both synchronous and asynchronous, between domain experts and computing specialists. By developing and disseminating open online resources by way of a curriculum, environment and illustrative materials for making construals, CONSTRUIT aims to promote making construals as a new digital skill developing sense-making, adaptation and repurposing skills that can provide a more appropriate foundation for the open education enterprise.
 CONSTRUIT project – http://www.construit.org/
 David Gooding, “Experiment and the Making of Meaning: Human Agency in Scientific Observation and Experiment”, Springer, 1990.