Over at ComputerWorld, I stumbled across this blog post “Iceland – Haven of Openness?” by open source evangelist, Glyn Moody. Now, I have to admit that I have never met the man and I’m sure that he’s a decent enough chap but I was left a little bemused by his article: I really have no idea what his point is intended to be. Which, for a journalist, ought to be worrying. The evidence of the eye ought to supplant the handwaving of the word. And the word in question is “open”.
Mr Moody seems comfortable in slipping between the terms “open source”, “open standards”, and “openness” with ease but without at any point really making any meaningful distinction between these very different concepts.
A major concern of mine is that the mantra “open” has become exactly that – a mantra, deliberately devoid of meaning, and used repeatedly by the speaker to achieve some transcendental state from which, supposedly, superior work is performed. Or, in more down to Earth terms, as my college politics professor used to say – it’s a “hoorah” word. You say “democracy”, “liberty”, “open” even, and the whole world shouts “hoorah!” – it makes you feel good but it still doesn’t mean anything without a context.
But it’s clever in a somewhat insidious way – after all, if you question “open”, you must be in favour of “closed”, no? You must belong to the dark side (Actually, let’s stick with that a moment – while “closed” might be a clear and unambiguous state, “open” is not. How open? Just a crack? Ajar? And what would “fully open” actually look like or mean?).
Mr Moody talks in his post about “slow progress in getting open source deployed by the UK Government”. Hmmm, strange axiom. I wouldn’t propose as a public sector IT strategy to see “progress in the deployment of Oracle databases” or “progress in the deployment of Microsoft Office”, so why on Earth should I single out open source as part of my strategy? My concern, surely, should be about progress in adopting and, when necessary, adapting, the technologies appropriate for the particular problems in hand. In seeing policy objectives met and giving useful guidance on how to meet them, not prescribing the details of “how”. My end game isn’t about the technology but about the service it delivers. To paraphrase Alan Cooper, talking about interface design, however beautiful your technology, I actually want to see less of it – it’s the results that count.
If Mr Moody expresses concern that there are fewer open source implementations in the public sector than he would like to see, maybe he should pause to consider if there might actually be less real demand than his handwaving and blind faith would have us believe. Or that open source, which has produced some arguably good stuff in some domains, sometimes just can’t deliver the innovative quality that comes with massive R&D budgets (think iPhone, Metro interface, Siri, etc.)? I can think of one massively successful open source project – Wikipedia – but have you ever tried to listen to a sound file on a Wikipedia page from a normally configured Mac or PC? I would argue that Wikipedia’s blind faith in open source has been a disservice to to the public. It has actually made an issue out of the software you run on your device and made that a more important issue than the content you are trying to access. That should never be so.
During my many years in the public sector (not, admittedly, in old Blighty but fairly representative as public sector organisations go), we had several major open source projects – some of which were real jewels and others which were sprawling messes. We had a few staff committed to open source, whatever the cost because they believed that “it is the right thing to do” (for want of a better description, let’s call this “faith-based” IT development), but for the most part, decisions on resource allocation came from level headed considerations about all aspects of the project: for example, the total cost of acquisition, ownership, maintenance and disposal of any particular application or system; or the ability of the particular technologies deployed to fit in within a wider ecosystem of disparate applications, loosely bound through open points of interoperability.
When I see the mess made in the German Foreign Ministry – fierce advocates of open source, if ever there were ones in the public sector – it gave me pause for thought: here was an important and major public sector department making an honest attempt to shift to open source in the belief that it would simplify their lives and make major savings for the taxpayers. After several years, they abandoned their attempts: they complained that even the different versions of open source office software couldn’t work with each other. After studying three scenarios (open source, proprietary, hybrid) and six criteria (security, cost, quality, user-friendliness, risk, and “political relevance”), they concluded that the open source scenario could not be realised and agreed to switch to PCs running Windows 7.
Was this a “disaster” for open source? Of course not. A valid bid was made against competing offers. You win some, you lose some, depending on the quality of your offer. That’s how competition works, provided that the playing field is level, so it would be strange that only one model – open source – were chosen up front as the way to go. Open source is not some neutral place from which public authorities should automatically make their IT procurement decisions. It is one business model among many, in which – like a lot of consulting work – the up front costs may seem lower compared with “off the shelf” solutions, but where costs are incurred and added all along the way. It’s not to say it’s wrong or better – it’s a different model and its pro’s and con’s should be weighed alongside any other business model.
If, however, you want to argue for open source as a matter of public policy, please feel free to do so, but don’t complain when there is public reaction against ideologically imposed solutions. This is not a rant against open source, but a rant against any attempt to dictate solutions to problems, or even worse – to claim that there is onlyever one solution to a given problem (which, I would hazard a guess, the open source guys – along with every proprietary supplier – would argue should be their solution). Most people do not care about the technology stack and are right not to, even if there is devil in the details. An analogy will hopefully enlighten: If you state as a matter of policy that you want to build a bridge, you will need to give clear indications such as: where from, where to, enviornmental and cost considerations. As a matter of policy you are not however going to specify that it is built of wood, stone, steel or according to a particular design. There are hundreds of ways of building bridges. Every successful one conforms with particular standards appropriate to the materials and design. There are good architects and engineers, as well as bad ones. It’s a market.
If there was truly only “one way” to build a bridge they would all look like stone arches with a keystone at the top! (In fact, the first ever iron bridge was built according to carpentry principles because “everyone knows” how you build a bridge from wood – standards evolved as understand of the materials evolved). If public authorities had always insisted that bridge building conformed with a single (set of) standards, imagine how poorly we would be served, without the Golden Gate, Øresund or Vasco de Gama bridges we have today. One standard does not fit all. And even if there were, who is to say that in technology, open source is the way (well, OK, an ideologue would – but that’s my point).
Another issue bugs me abut Mr Moody’s piece: the interchange – casual, deliberate or otherwise – between two distinct terms “open source” and “open standards” and the resulting confusion. Whilst open standards may have emerged over the last decade inspired by a community-driven development model that is similar to that of the open source movement, open standards and open source are two orthogonal issues.
Open source refelcts a wide umbrella of intellectual property models for software development. Open standards are policy products coming out of a range of organisations around the world and against which any of a range of software development models can build implementations, including open source. A good open standard will specify clear conformance criteria and requirements without reference to any technology stack or at least distinguish between the ‘core’ standard and technology-specific ‘profiles’. This has become an industry best practice.
The different processes for developing open standards are careful about how they prescribe specific intellectual property modes. Many organisations develop standards according to one or other specific IP modes, whether that be FRAND, RAND, Royalty-Free, Non-Assert Covenant, etc. Some organisations have a range of modes that participants can choose between. Whether a standard is considered “open” or not is not, and should not, be a reflection of the intellectual property possibly associated with it. Is MPEG a set of open standards? Bluetooth? WiFi? ODF? They all contain IP in one form or another – what is relevant to policy makers should be the process by which these standards are developed and maintained and the quality of the standard as a “product”. As in the bridge-building, it is the solution which is important, not the materials. Various standards help assess the pro’s and con’s of different approaches and solutions.
Rolling together “open source” and “open standards” in the same breath serves only no-one, except maybe a few who would wish, surreptitiously, to introduce and benefit from a specific business model – open source – on the coattails of a laudable process – open standards. Classic “bait and switch”. Caveat emptor. I’m not sure where Mr Moody stands on this. His credentials suggest that he is a disinterested observer and analyst. He is clearly also an ideologue and passionate open source advocate and I wish him the best of British. But, please, don’t bang on about “open standards” when really concerned with promoting open source. It does everyone a disservice.