Yesterday I had another meeting with Richard Steel, CIO of Newham council, who was generous with both his time and information. After our introductory session a few weeks ago, we got down to the nitty-gritty – the server side.
I was impressed by spaghetti-like complexity of the diagram showing the links between the disparate services and their databases: running a borough is clearly an incredibly complex job, and it's clear that we are still in the early days of automating that process.
Two things emerged during the morning, one good, and one bad.
First the bad. Newham demonstrates that once Microsoft products are used for a large number of functions in a large organisation, there is a natural tendency to use even more of them because of the way that Microsoft links and binds them together. As more and more Microsoft-based skill sets are acquired, the switching costs become very high – which is precisely why Microsoft adopts this tightly-integrated approach.
This means that, realistically, there is little scope for swapping in open source solutions to replace those of Microsoft, even when the total cost of acquiring and running the software is lower. The re-training costs will always be a barrier.
But all is not lost. Another thing that emerged from the meeting was that Newham – and other councils – do write their own software for certain more specialised areas. Since many functions are common to all councils, there is clearly scope for adopting an open source approach here: turning the code into a commons would allow the councils to collaborate on it, reducing the cost for each of them. It would also enable them to develop the code faster.
This kind of pooling probably requires an impetus from central government, since each council naturally tends to worry about its local needs first.
Indeed, that was perhaps the most important insight that I gained from yesterday's meeting: that local councils find themselves in something of a Prisoner's Dilemma when it comes to choosing whether to go with Microsoft or free software.
Individually, it makes sense to do deals with Microsoft, since councils can use the threat of turning to open source to obtain better deals. But if they *all* turned to open source, the overall cost savings would be much greater.
Finding staff with the relevant skills would be easier, since all councils would be developing in the same way. This would help ease – although not totally solve - the switching cost problem, since all councils would be training staff in the use of open source.
If a council switched with the present situation, it would be faced with the need constantly to retrain staff brought up on Microsoft solutions as they joined. But if open source competence were spread around the country, there would only be one round of such training: thereafter, people moving between councils would bring the right skills with them.