UK Government Consultation on Open Standards

During UKGovCamp in January, three of the people most closely associated with issues around IT standards in UK Government (James Forrester, Hadley Beeman and Liam Maxwell) gave us a preview of a consultation that was due to open in February. I remember them jumping through all sorts of hoops so as not to say more than they were allowed to as the consultation hadn't yet been formally launched (hard for the three of them, fun for the rest of us to watch). I made a mental note to make sure I knew when it was opened (I followed @liammax) and when it did, let the relevant folk at W3C know about it. I don't think I'm breaking any confidence to say that we thought "yeah we'll submit some answers – we've got most of them somewhere already, we do this all the time." In other words, I signally failed to predict quite how important this was going to be to W3C and how much of my time it was going to take up, mainly because I live in a world where to say that royalty free open standards are a good thing is to state the bleedin' obvious.

What's the consultation about? HMG is proposing adopting policy whereby any IT procurement contract would need to specify the use of open standards for software interoperability, document and data formats. If it didn't, the relevant department would need to explain, publicly, why open standards were not specified (this is the comply or explain principle). The policy may or may not allow government departments to mandate specific open standards. This raises several issues:

What is an open standard?
Quick answer – one developed through an open process (and the W3C process is often held up as the prime example of how to do it).
Does the use of open standards for software interoperability, document and data formats create a level playing field for SMEs and enterprise alike when they compete for government IT contracts? (that's the intention but does it?)
Quick answer – yes of course it does, do you really have to ask?
Does "open standard" imply royalty free or is (F)RAND equally acceptable, particularly for open source software?
Quick answer: no, not all open standards are royalty free, the two are separate, however, open standards for software interoperability, document and data formats are royalty free and that's the best way to level the playing field.

Fast forward to April.

The responses were not as expected. Actually, they were very heavily against the policy. Microsoft in particular has been making strenuous efforts to lobby against the policy. I can say that openly and without caveat as it is clearly true. Witness the lobbying uncovered by Glyn Moody and reported in a series of articles beginning with this one concerning a previous (related) consultation and the débâcle over Andy Hopkirk's involvement as both facilitator for one of the early events and adviser to Microsoft that has lead to that event being struck from the record and the consultation being extended by another month. Microsoft is not against open standards per se — they contribute to their creation and implement them massively — but there are areas where their interests might be at risk from the proposed policy and they are therefore being cautious. In particular, they're keen to establish where, if anywhere, their business would be affected.

Standards organisations like W3C, Open Source advocates, SMEs and enterprise that use open source software and whose business depends on open standards needed to up their game. And we have — in spades.

The most recent formal event in the consultation is well-described by Mark Ballard's article Software industry reclaims open standards debate. As Mark pointed out, the input from Iain Mitchell QC was authoritative and illuminating. He referred in his presentation to a Legal Opinion he'd recently prepared: Compatibility Of The Licensing Of Embedded Patents With Open Source Licensing Terms which I can thoroughly recommend (it is very readable). The killer input for me though came from Simon Phipps. He, Linda Humphries and Liam Maxell were able to point out that when Keith Mallinson said things like "most standards are published under FRAND conditions" he was, like all of us do all the time, speaking from his experience. And his experience is in a world where this is true (telecoms). It just happens to be irrelevant to the issue of software interoperability, document and data formats in government IT. For bodies like the British Standards Institute and the Object Management Group it's royalty free all the way with a trivial number of exceptions. For W3C there are no exceptions. Simon Phipps' article Accommodating Telecoms explains this very well and, as I tweeted when I saw it, I think it is possibly the single most useful contribution to the debate. Basically it cuts the crap and gets to the point.

My role has been to present the W3C case which I have done now at three events. I've also worked on our formal response with various colleagues (it's not public yet but soon will be, it's going through one more round of internal checks). Speaking on behalf of W3C is scary. It's a big brand that carries a lot of weight and one that I care deeply about. What if I say the wrong thing? What if I get cornered into saying something I shouldn't? Thankfully I was very well briefed by W3C's counsel Rigo Wenning who gave the sage advice of "stick to talking about the Web and you won't go wrong."

The formal events organised by the Cabinet Office are centred on specific questions within the consultation. Each person is asked to provide answers to three questions and then they're discussed. That allows Linda Humphries, the person actually collating the responses, to gather direct input in a structured way. But it doesn't allow for what one might call 'free text.' I got a chance to do that at an event organised by the Policy Exchange on 30 April: Open standards for open government?

Here's the argument in brief.

In March 1989 Tim Berners-Lee submitted his paper Information Management – A Proposal. It was his solution to the problem of software interoperability, data and document formats. Different labs and research teams at CERN used different systems and could barely speak to each otherr. It was, to use the mot du jour, an Omnishambles*. Tim's solution was to see if he could get people to agree on a few simple things and from that came the World Wide Web.

The success of the Web comes, in part, from the fact that the standards on which it is based only cover areas of essential commonality with minimal restrictions on specific applications. Take just three core standards:

  1. HTML – not just a document format but an interactive platform that underpins the biggest revolution in publishing since Guttenberg;
  2. XML – the massively implemented exchange format for structured data;
  3. RDF – the way to publish data at Web scale.

All these technologies underpin a stack of technologies and toolkits, many, but not all, of which are open source.

For software interoperability, document and data formats, the Web proves the value of royalty free open standards. For governments to exchange information between different departments and, even more importantly, between itself and its citizens, there simply is no better solution. Furthermore, the Web works on desktops and laptops, yes, but it's also available on mobile phones, games consoles and Web-enabled TVs. Surely systems based on royalty free open standards must be a big part of efforts to minimise the digital divide.

* I meant to use this word in my talk but forgot and I've been kicking myself ever since. It was coined by Armando Iannucci in his political satire The Thick Of it and (Shadow Home Secretary) Yvette Cooper used it recently in the House of Commons to describe the current government. The relevant clip is, of course, on YouTube (NSFW).