Loosely Coupled v Tightly Integrated

I’ve been mulling over the loosely coupled versus tightly integrated systems development arguments and in particular Martin Weller has a number of posts on the subject including this one. In it talks about the control issues which are faced by many organisations. I’m in agreement with him about favouring a less restrictive approach but perhaps not for the reasons he gives. My main reason for having a set of ‘loosly coupled’ services many of which may be externally provided is for the following reasons :-

1. Expertise through experience – People have spent years developing specific tools to serve particular purposes, they do this very well (which is why their products have lasted years) so they give a rich and fulfilling user experience. When these are re-created within a VLE much of the richness appears to be lost. For an example of this I’ve always used FirstClass which is a CMC system (plus some other bits as well now) but at it’s heart is a well designed interface of conferencing. When we move to a VLE the alternative provided is a forum posting system. This reduces the functionality and user experience significantly just so that it can be ‘tightly integrated’.

2. Complex systems lead to complex solutions – When you try to develop something to do everything for your business (Enterprise Content Management and Virtual Learning Environments being two that we are grappling with at the moment) they rarely knock down all the pins. It’s likely they’ll do a lot of good things but that shouldn’t then mean they replace everything that went before them. If they don’t do something as well as it can be done externally (for example blogging) then organisations should look to use an external blogging service instead. Organisations should get over the ‘not invented here’ or even the ‘not commissioned here’ syndrome.

3. Your staff and students use other systems anyhow – In staff and students are using external blogs, wikis, social networks, search engines and so forth so the idea of controlling those people to use your own in-house system is mute. Once you recognise this fact it allows you to think in a more loosely connected fashion about how to join up resources and tools. Things such as Google Desktop start to become much more of a business tool. Read my previous posts about horizon scanning for some information about how businesses are picking up on this idea of incorporating some of these tools.

4. Get someone else to do the work – Why build and support a product when someone else can do it for you? All you need to do it integrate it. If new browsers arrive then it’s someone else’s problem to ensure that the products work with them. You can of course help (in an open and collaborative fashion) but the effort is hugely reduced by not building everything in-house.

5. People like to share – There it is. In organisations where information is tightly controlled there are always communications difficulties and duplications taking place. I don’t mean that we should open everything up to the world (‘going naked’ as people keep saying) but transparency is no bad thing. Where it can be applied, it should be applied.


About willwoods
I'm Head of Learning and Teaching Technologies in the Institute of Educational Technology at the Open University.

2 Responses to Loosely Coupled v Tightly Integrated

  1. ajcann says:

    For me No. 3 is the killer argument. You can’t control user/student behavior, and if you try too hard, you’ll only drive people away. This prescriptive stems from the days when universities had the lockdown on emerging technologies. But universities are conservative institutions and it’s hard to change decades of embedded practice.

  2. willwoods says:

    I agree. It’s also true that University central IT departments see remote hosting as inherently insecure. Although strangely I notice that according to IT week businesses are flocking to remotely hosted solutions.

    Universities’ operational services seem to lag behind other sectors in changing to meet new ways of working even though we lead on the research. Is that true everwhere I wonder? – MIT seem to manage to squeeze stuff through from research to delivery relatively quickly (e.g. the Secondlife courses). Are there models we can adopt in the UK? – I know Cambridge is working with MIT on a number of projects some of which I’ve been involved in however even they don’t seem to do the conversion through to delivery as quickly and easily as in the U.S. – Is that just my perception or is it a truism?

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: