Open Source – who poached whose policy?

Only a couple of weeks after the Tories announced a policy to support Open Source, yesterday the Government published a new policy on Open Source software “that will ensure maximum value for money for taxpayers”.

Unlike the Tories though, the Government has put some meat on the bones of its press release and published 10 actions that “will actively help make sure the best possible, best value for money software solutions are put forward for tenders, be they Open Source or propriety products”. However, as with the Tory announcement, I believe the Government policy is fatally flawed.

Fortunately, those 10 policies/actions recognise the need to make procurement decisions “on the basis on the best value for money solution to the business requirement, taking account of total lifetime cost of ownership of the solution”. However, the policies/actions propose to help tip the balance towards Open Source through the introduction of two proposed safeguards that are, in my mind, unworkable in practice:

Firstly, where Government becomes locked into proprietary software it will, where possible “take exit, rebid and rebuild costs into account in procurement decisions and will require those proposing proprietary software to specify how exit would be achieved”. Whilst proprietary software suppliers can no doubt describe the exit route, it is impractical for either the supplier or the purchaser to try to estimate rebid and rebuild costs with any degree of accuracy– and who would be responsible for such costs if they were subsequently (say, 5 years down the line) proven to be wrong?

However, and rather worryingly, it does introduce the opportunity for customers who have already decided on their preferred solution to justify its purchase over, say a solution with a lower overall cost of ownership, by coming up with an over-stated exit, rebid and rebuild cost.

Secondly, “the Government will look to secure full rights to bespoke software code or customisations of commercial off the shelf products it procures, so as to enable straightforward re-use elsewhere in the public sector.” This is fine for bespoke software that can work in isolation, but is virtually unworkable for bespoke code and customisations that rely on products where the IPR is not owned by Government.

Also, this approach could potentially increase maintenance costs for Government. Within my company, Radius our normal approach for making most customer-funded bespoke modifications to our own products was to retain the IPR, but roll the modification into the standard product and make it available to all other users (Public Sector or not). This way the modification was taken forward in new releases, and maintained, under the standard maintenance agreement, at no additional charge.

If the IPR is retained by Government, then the code cannot be rolled into the standard product (i.e. to be made available to others outside the Public Sector), and presumably the Government will be happy to pay for its maintenance and inclusion in all future releases of the standard software....

I’m also bemused by the statement – “The Government will use open standards in its procurement specifications and require solutions to comply with open standards. The Government will support the development of open standards and specifications.” I can’t think how many times I’ve seen this statement over the past 25 years, be it for mainframe, Unix , e-government or general interoperability – great in theory, but never been implemented effectively in practice.

I won’t repeat the arguments contained in my post Conservative’s move to Open Source – a major mistake?, other than restating that the trick is to learn to live with the proprietary software suppliers (accepting that the UK Government software market is a significant, but still very small percentage of their overall business) rather than build ways to avoid them.