This page is based on a presentation given to the Ubuntu in Business event and draws on experiences of an open source strategy in a 400 person, 5 office Anglo-French heritage services company. Those who think that schools are fundamentally different to other organisations in the way they should implement , manage and strategise around technology should maybe stop here. Everyone else, pitch in!

Suggested approach: this page stands as a summary, bare-bones framework. I've tried to cast my mind back to when I worked in schools and slant the summary towards school's interests. Each element should eventually link to its own page with further explanation, examples to help see how to advance that element, etc., but the framework summary should be sufficient for SMTs and governing bodies to start the ball rolling.

The framework is presented as a series of steps, although it is clear that some steps can and even should be taken in parallel with other steps. The general principle that the lower numbered steps need to happen before the later ones is however one to be guided by. The framework might be subtitled "Making your school better for less" .

Step 0: Acceptance that open is inevitable, because, among other reasons, it is the best strategy for increasing inclusion, flexibility, sustainability and reducing costs. Adoption of an open source strategy as a school or group of schools ensures every child can have access to the software used by the school, allows every school to benefit from the resources developed around the software without any additional cost and puts the school on to an upgrade path that does not come with an annual price tag. Longer term an open source strategy that has been adopted by 10s of schools and education organisations will inevitably result in better software for education.

Step 1: Identify core services delivered by technologies with high dependencies. Look at software running on your servers and work out which has the greatest dependences aka the least numbers of choices for both the server operating system and full access by any related client applications.

Step 2: Identify replacements that, at worst, have lower dependencies than those they replace and at best are open source with multiple organisations offering commercial support. This step can require a long term outlook; some applications will not necessarily have replacements readily available; it can also require a little thinking outside of the box, going back to first principles; maybe to achieve the goal a like-for -like replacement is not necessary nor even helpful.

Step 3: One by one, replace the server operating systems with Ubuntu (or another well-supported Linux distribution), KVM (a well-supported open source virtual server platform). This step is necessarily in parallel with both step 2 and 4.

Step 4: Install replacement core services; understandably in parallel with step 3, unless you want to opt for a "big bang" step change over the summer break. This latter approach not something that is generally advisable unless you have access to a senior and very strong IS Management professional and a committed SMT!

Step 5: Identify desktop applications with high dependencies. Similarly to step 1, look at software running on your desktops and work out which has the greatest dependencies, in particular those which cannot run on common Linux desktops. Focus on the most widely used applications; those with limited use might have to stay or be accessed via remote desktop solutions, and it is also the most widely used that have the greatest cost, sustainability and inclusion implications.

Step 6: Find replacements. This is in many ways easier than the server equivalent. Firefox or Chrome are better web browsers, OpenOffice is more than enough for the great majority of users, etc..

Step 7: Persuade people that feature parity is not essential, and probably irrelevant; that good enough is good enough. This can be quite challenging as 20+ years of software sales/marketing has engrained in our psyches that more features=better, even when we didn't use most of the features available in the old version. Of course, open alternatives are often far more than good enough and have a lot of features; arguably some parts of education could do with cut-down versions. This is of course an option with open source software!

Step 8: If only just because you can, install the replacements everywhere (that they will do no harm). Flexibility, ease and reduced cost of rollout, ...

Step 9: Remove old desktop applications

Step 10: Replace desktop OS with modern shiny open one (that would be Kubuntu, of course)