Puffbox

With some unexpected free time in my schedule for today, I sat down to complete the Cabinet Office’s recently published questionnaire on Open Standards in government. To be perfectly frank, its 120 questions have left me reeling, and nervous.

Screen after screen of acronyms and document reference numbers – ISO this, BSI that, W3C whatever – which you’re asked to rank as priority, mandatory, recommended, optional, or ‘should not use’. They swing from the insanely detailed to the laughably obvious. At one extreme: I can’t believe more than a handful of people on the planet understand the finer points of ISO/TS 16071:2003 as opposed to ISO 9241-171:2008… and I’m not entirely sure they would be the right people to be making decisions affecting day-to-day hands-on use anyway. At the other: yes, I think JPGs are probably a good image format to use.

The thing about standards is, even a bad standard is a good standard.

OK, I’m over-simplifying a little. There might be certain reasons why one particular refinement of XML is better than another for a specific purpose. But broadly speaking, as long as you’re giving me XML, I’m sure I’ll be able to deal with it.

I don’t think we have the money to spend on librarians and uber-consultants, the only people who’ll really know or care about this stuff, holding talking shops about which particular ISO standard is just right. It’s the absolute opposite of the ‘Agile’ philosophy I thought we were all supposed to be moving to? And what if The Market decides that it prefers a different standard… or more likely, that it just doesn’t care? It should be perfectly possible to deliver well-structured data in whatever format people may require: be this, to take one specific web-related example, HTML4 or XHTML or HTML5 or RSS 1.0 or RSS 2.0 or Atom or NewsML.

I can’t help feeling that we’re coming at this from the wrong angle altogether. The problem isn’t with the selection of a particular open standard. The problem comes when government chooses closed standards (er, MS Office?)… or even worse, decides to create standards of its own (eg IPSV).

The end game on this, I’m convinced, is the centralised provision of open-source-based platforms, which are – by their very nature – standardised and open. Imagine if we had all government news output in a single multisite WordPress instance, and someone somewhere asked to receive material in NewsML. WordPress does a lot of formats ‘out of the box’, but not NewsML. However, WordPress is built with the expectation that you’ll want to add to it. It has an add_feed function, which would let you create a new feed output template, call out whatever data you needed, and drop it into the right place. A few days work to code a plugin, activate it network-wide, and you’re done.

So Cabinet Office, I’ve answered your questions. You’ll end up with a piece of paper, ranking other pieces of paper in priority order. Will it answer the question? I fear not.

Oh, and just for the record: neither the SurveyMonkey questionnaire, nor the Cabinet Office website’s page pointing to it, are [at the time I write this] compliant with the 11 year old XHTML 1.0 Transitional standard they both declare, according to the W3C’s validator. I know it’s a cheap shot, but it puts things in some kind of context.