Puffbox

Simon Dickson's gov-tech blog, active 2005-14. Because permalinks.

2014 | 2013 | 2012 | 2011 | 2010 | 2009 | 2008 | 2007 | 2006 | 2005

Code For The People company e-government news politics technology Uncategorised

api award barackobama barcampukgovweb bbc bis blogging blogs bonanza borisjohnson branding broaderbenefits buddypress budget cabinetoffice careandsupport chrischant civilservice coi commentariat commons conservatives consultation coveritlive crimemapping dailymail datasharing datastandards davidcameron defra democracy dfid directgov dius downingstreet drupal engagement facebook flickr foi foreignoffice francismaude freedata gds google gordonbrown governanceofbritain govuk guardian guidofawkes health hosting innovation internetexplorer labourparty libdems liveblog lynnefeatherstone maps marthalanefox mashup microsoft MPs mysociety nhs onepolitics opensource ordnancesurvey ournhs parliament petitions politics powerofinformation pressoffice puffbox rationalisation reshuffle rss simonwheatley skunkworks skynews statistics stephenhale stephgray telegraph toldyouso tomloosemore tomwatson transparency transport treasury twitter typepad video walesoffice wordcamp wordcampuk wordpress wordupwhitehall youtube

Privacy Policy

  • Geolocation: getting worse?

    Credit: Roo Reynolds

    About a year ago, I ran a short experiment, with the assistance of the various misguided fools who choose to follow me on Twitter. I was intrigued by the possibilities offered by the geolocation element within Google’s javascript API. With one line of code, you could theoretically make a good guess as to where the user was physically located – and could tailor the page content accordingly. But did the theory hold up?

    The answer – sadly – was no, not really. When it worked, it was brilliant; but all too often, it didn’t work. Some ISPs reported curious results; sometimes it gave an error; some ISPs seemed to block it outright. So my conclusion then was that ‘it’s not really good enough to make meaningful use of’.

    One year later, and I’m working on a few websites for MPs/candidates ahead of the general election. One specific project involves a front-bencher with interests at both local and national level. It would be great if we could show local-level things to people in the constituency, and national-level things to people outside. Had the Google API improved at all?

    Sadly – no. In fact, if anything, it had got worse. Mobile seems to be a particular blackspot: nobody using a phone reported a success. And the various changes in the ISP market don’t seem to have helped. We had people in Leeds and Manchester being told they were in London, and people in Reading being told they were in Glasgow.

    So I’m reluctantly abandoning the work I’d done on automatic location detection. Instead, we’ll show you one view (either ‘local’ or ‘national’) by default, with an obvious method to switch to the other; and will drop a cookie each time you switch, so you’ll see the same view when you return next time.  Not a bad Plan B, really.

    There’s no doubt in my mind, ‘proper’ geolocation is going to be a very big deal – but for the foreseeable future it’ll be dependent on people running the most modern browsing technology. That’s fine for mobiles; but we live in a world where IE6 still refuses to go away. (Actually… according to this source, IE6 is the single most popular browser version out there?!)

    Picture: Roo Reynolds, Flickr

  • ,

    Commercial advertising on govt's no1 website

    Figures from Hitwise at the end of November showed the Met Office to be central government’s most popular website, with traffic getting on for double that of Directgov. Now I must admit, I’m not a regular visitor to the site: I tend to favour Metcheck. But with the country apparently heading for a new ice age, I thought I’d drop by, and see their prediction for the days ahead. And I got quite a surprise:

    But that’s not all. Every few refreshes, you’ll also see one of these:

    I seem to recall they’ve had ‘house adverts’ for their own commercial products for some time. But this is the first time I’ve been aware of third-party commercial advertising on the site.

    I’m not saying it’s a bad thing. They’ve clearly got healthy traffic levels, and they’re a Trading Fund: if advertisers want to hand over cash for a few pixels of unused screen space, hey – shouldn’t they be duty-bound to accept it? (There’s more information in this press release from September… including the rather startling suggestion that their traffic peaks at 19 million visits per day.)

  • ,
    , , ,

    Wanted: consultation platform, £1m reward

    I’m glad my former Microsoft colleague John McGarvey reminded me of Conservative shadow culture secretary Jeremy Hunt’s proposal of a £1m prize to develop ‘the best new technology platform that helps people come together to solve the problems that matter to them’. That’s what happens when you announce things over the Christmas holidays.

    The plan is for a future Conservative government to use it ‘to throw open the policy making process to the public, and harness the wisdom of the crowd so that the public can collaborate to improve government policy. For example, a Conservative government would publish all government Green Papers on this platform, so that everyone can have their say on government policies, and feed in their ideas to make them better.’ Why does that sound so familiar? ‘There are currently no technological platforms that enable in-depth online collaboration on the scale required by Government,’ says Mr Hunt; ‘this prize is a good and cost-effective way of getting one.’

    Now I don’t know what kind of ‘scale’ or ‘depth’ Mr Hunt thinks he requires. If there’s a formal brief, I’ve yet to find it – and I’d be delighted if someone could point me in the right direction.

    Because I’ve been building websites allowing the public to input their views on government green and white papers for some time now. Steph Gray’s Commentariat theme kickstarted the process: and I’ve since gone on to build reusable WordPress MU-based platforms for two Whitehall departments, for a few grand each. We’ve proven WordPress can handle (literally) thousands of responses – and in the only case so far where it’s wobbled, that was because of ISP throttling rather than the ability of WordPress to handle it.

    Then on the academic side, you’ve got the work that’s been done by Joss Winn and Tony Hirst et al on JISCPress / digress.it / writetoreply.org. Their focus has been on the technical side, including some early steps towards community-building. It’s a bit lacking in terms of aesthetics, and it hasn’t yet been tested with huge volumes, but it’s doing some very interesting things.

    And of course, barely a month ago, you had Mr Hunt’s own people at Tory central office proving the point by turning the government’s draft IT strategy into a consultation document using WordPress. Cheap and quick, showing signs of inexperience with the platform – but good enough to receive nearly 400 contributions.

    So you have several independent operations in the (wide) UK public sector, already proving in the real world that WordPress is perfectly capable of supporting such ‘user feedback’ websites, and delivering some pretty sophisticated functionality and user experience. BuddyPress, meanwhile, continues to improve, and could certainly form the bedrock of a government-backed policy development community.

    There’s no doubt in my mind that the technology is ready. And there are enough good people who have built up enough experience to collaborate on building something pretty special. For a slice of that £1m, I’m sure I could find time in my own schedule.

    But the big question is… is Mr Hunt ready? What does it mean to receive large volumes of contributions from the general public? When do you ask for them? How do you deal with them? How do you ensure they’re representative? And what if you don’t like the consensus of the opinions expressed?

    I’m all for the kind of revolution in policy development he seems to be proposing; and I’d be happy to play a part in it. But it isn’t the lack of a technical platform that will hold this vision back. If anything, that’s the easiest part.

    PS Just a thought… whither Tom Steinberg?

  • 2009: the year I became a developer (sort of)

    When I started in this business, I made a conscious decision not to become a programmer. I knew I had it in me: if I could crack Latin, I could certainly crack PHP. But I’ve always recognised that I’m better ‘across the board’ than most people I come across in the field. There just aren’t that many people who can appreciate design and development and editorial and Westminster. And besides, if I decided I wanted to be a developer, I’d have to concentrate 100% on it.

    And yet somehow, at some point during the summer of 2009, I started cranking out more and more ambitious code. My PHP efforts went beyond straightforward HTML templates with WordPress tags dropped in. I wasn’t scared to look at javascript. Next thing I know, I’m writing WordPress plugins and pretty advanced javascript/Ajax routines. I’m scraping web pages in their thousands, to get data in the form I want. All stuff I knew was possible, and probably understood on a superficial level – but here I am, doing it. Dammit. So how on earth did I get here?

    A lot of it is down to WordPress, which acted as a gateway into the depths of PHP. You can achieve a heck of a lot in WordPress with fairly sketchy PHP knowledge – following the Codex‘s instructions, and not asking too many questions. But inevitably I found myself wanting to dig a little deeper: to understand why certain things did what they did, and to find out what other options were available. I realised I’d unconsciously picked up quite a lot of the basics, enough to understand the more complex concepts.

    It’s also been the availability and maturity of certain tools: in particular JQuery and SimplePie. The former is the perfect route into javascript, making pretty advanced techniques seem as straightforward as CSS. The latter makes it laughably easy to work with RSS – opening the doors to all sorts of possibilities, where feeds are available. It’s also been extremely helpful to find a couple of CSS frameworks I’m comfortable with, namely YUI and 960.gs – simplifying the layout process and letting me devote my time to other aspects of the work.

    Part of it, too, has been the challenges thrown up by several key projects. For example, there was a moment in the Lynne Featherstone project where I discovered an unexpectedly huge amount of unstructured HTML content to be imported. I’d never looked at screen-scraping in much detail: sure, I’d played around with it, but I’d never yet had a reason to get my hands seriously dirty with it. To my great relief, I came up with a (so far) reliable method for scraping entire websites into a format suitable for WordPress import… and I’ve had cause to use it on a couple of other projects already. Necessity, being the mother of invention, has added several such strings to my bow.

    And I have to say, meeting some great people through the year – particularly within the WordPress community – has been a further encouragement. It turns out, developers can be relatively nice, relatively normal guys. Blimey, one or two of them might even qualify as cool.

    I’m not ‘one of them’ yet, nor do I ultimately want to be. I’m still in awe of, and slightly intimidated by, the really good ones. I’m sure a real developer would look at some of my work, and laugh. But by and large, my stuff works as well as it has to work. And even if it doesn’t, we can call it a prototype until someone more skilled can come along and do it properly. 🙂

  • ,

    The puzzle of political donations

    Just over a year ago, we had the startling success of the Atheist Bus Campaign, which raised over £150,000 to put the message ‘There’s probably no God, now stop worrying and enjoy your life‘ on the side of London buses. And, given the surplus funds, buses up and down Great Britain. An online-orchestrated grassroots campaign, with individuals chipping in a few quid for a political gesture.

    Fast-forward to this weekend, and the least festive Christmas Number One Single ever, with Rage Against The Machine’s foul-mouthed tirade seeing off a sweet Geordie teenager – solely on the back of download sales. Again, a large number of people, orchestrated online, chipping in their 29-99p for a political gesture. Of sorts. Well, political in the widest possible sense.

    I’m talking to a few MPs at the moment about improving their websites ahead of next year’s election, and one recurring subject is political donations. On the face of it, they’re being most optimistic. The biggest news story of the year has been the way MPs have fleeced the public purse to fund their moat-cleaning and duck-houses. I don’t see queues of people outside constituency offices, all eager to hand in their campaign contributions.

    But it seems it can work. Iain Dale observed at the weekend that Tory candidate for Bristol East Adeela Shafi is successfully raising funds through the MyConservatives.com site: over £1,000 towards her notional target of £1,500, apparently within 3 days. How? Iain suggests:

    I suspect most of the donations were in the £5-£20 bracket. If you make it easy to donate, people will do so, if you give them a reason to.

    …which is probably true; but he also hints at the significance of #kerryout, a Twitter-orchestrated attempt to unsettle, if not unseat, Labour’s ‘Twitter tsar‘ Kerry McCarthy. It’s supposedly independent; but that’s somewhat hard to justify when its hastily produced website includes a huge link to Adeela Shafi’s MyConservatives page for donations. And I wonder how the concealment of the domain name’s ownership sits with election imprint rules.

    So anyway, what can we learn from this?

    Getting detailed data on RATM’s sales breakdown or MyConservatives donations wasn’t going to be easy; but helpfully, virtually all the donations to the Atheist Bus Campaign were listed on the JustGiving website: so I did some hasty number-crunching. And here’s what I found:

    • The campaign raised just over £148,000 from 9,744 individual online donations.
    • Two £3,000 donations accounted for 4% of the total; 12 of £1,000+ accounted for 14.4%.
    • 162 donations of £100+ – or 1.7% of the donations – accounted for 30.8% of the total sum raised.
    • Half the money was raised by just 10% of the donors.
    • Roughly 76.6% of the donations were in Iain’s £5-20 range: but they accounted for less than half (49.7%) of the amount raised.
    • 14.5% of donations were £5 or less, making up just 10.6% of the total raised.
    • 81% of donations were £10 or less, representing 35% of the total contributed.
    • The mean donation was £15.19; the mode and median were both £10.
    • 36% gave exactly £10; 26% gave £5; 12% gave the minimum £2; 11.5% gave exactly £20.
    • 49 people gave exactly £6.66. Very amusing.

    So that’s a very small number of very large contributions representing a high proportion of the total raised; but still, over a third – that’s well over £50,000 – coming from donations of £10 or less. (Not that it’s any kind of sensible comparison: but the Obama campaign raised 38% of its funds from donations of $200 or less; the Atheist Bus raised 78%.)

    But for me, the two uniting factors across these three success stories are as follows:

    • they were negative campaigns – in the sense that they were based around someone or something that people didn’t like: religious advertising, Simon Cowell, Kerry McCarthy; and
    • there was a specific, measurable outcome: the sight of a bus with a poster on it, the announcement of the Christmas chart, the result from Bristol East on election night. If enough of you support me, we will get ‘X‘ – and we will know if/when we have won.

    If you were hoping for any tangible conclusions, I’m sorry to disappoint. But there’s definitely food for thought in there.

  • ,
    , ,

    Wireframes? Specs? Ha.

    I’ve added a lengthy comment to Stephen Hale’s recent blog post about preparations for a much-needed redesign of the FCO’s blogs.fco.gov.uk site. Unfortunately, the FCO’s platform did horrible things to the formatting, so even if it’s only to make it legible, I thought I’d echo one of the more controversial points I made in that comment.

    Specifically: my point that, for a project like that, the days of spending weeks and months honing wireframe diagrams and/or lengthy functional specifications should be behind us.

    A blog platform is no longer a start-from-scratch, blank-sheet-of-paper kind of project. Wipe away the surface layer, and there’s a very limited range of web page layouts these days. The functionality of a blog platform is even more standardised, with only a handful of serious candidates. Virtually all the functionality you’ll need will be ready, out of the box, within a matter of minutes.

    Having done this very regularly for several years now, I strongly believe that if you have a fairly good idea of the functionality you want, and a fairly good idea of the platform you like, you should look to force the two together at the earliest possible opportunity, rather than spending ages and £££ refining your wireframes and technical spec to perfection. Why waste time and money dreaming of what you might like, when you can have it in front of you within minutes, and know?

    It’s like when you buy a new car. Cars are a mature technology. They all feel a bit different, and come with slightly different features, but they all do broadly the same thing in the same way. If you want a new car, you don’t sit down and design your dream car. You don’t recruit your own team of engineers, designers and mechanics. You make a list of the few things that are important to you; then you go to the local showrooms and test-drive a few.

    In writing my comment for the FCO site, I went out of my way to avoid using the word WordPress. But my blog, my rules. So here’s the slightly less diplomatic version of what I wanted to say.

    • In a world of instant zero-cost availability, it’s ludicrous to consider functionality and platform in complete isolation from each other. It just is.
    • WordPress’s status as the world’s leading blogging platform is now, I’d suggest, undisputed. So if you want to run a multi-author blogging arrangement, it should be on WordPress. If you don’t believe me, maybe you could ask the Telegraph: they tried a bespoke platform, then tried a commercial product, then finally saw sense.
    • DFID are already running a multi-blogger platform, based on WordPress, and have been doing so most successfully for the last 15 months. It can do everything that you’d expect any such site to do – and more. It’s unquestionably a better system than the FCO’s. It ticks all the boxes on the FCO’s future wireframes; and if there’s anything it can’t already do, it can almost certainly be grafted on: that’s the beauty of WordPress. And we’ve proven that with them numerous times.
    • The DFID code is open source. Some of the key plugins are already available to the world on wordpress.org; I’m happy to explain and share any lower-level stuff within the templates.

    So…

    • If FCO come up with a reason why they can’t use the world-leading and lowest-cost solution, in conjunction with code already proven within government and also freely available, I sincerely look forward to hearing it. And I imagine Parliament will too.
  • ,

    DCSF joins WordPress trend

    DCSFchildrenplanIt’s now two years since DCSF published their Children’s Plan – I know – and Ed Balls wants to know what impact it has had on you. They’ve published a progress report, and launched a commentable website… based on, guess what, WordPress. Not the first time they’ve gone down the open source route: a year ago they launched their National Strategies website on Drupal. But I think this is their first WordPress-based site.

    Hold on a second though. What’s going on with that URL: gscdevelopment.com/wpsample? Well, gscdevelopment.com is simply an account at Bluehost – a very low-cost shared hosting provider. Nothing wrong with that at all; I use them myself for experimental hosting space, although I’m not sure I’d host a government site there. Google has literally nothing about a web development agency called GSC Development. And that’s a bit of a problem. It may look like a DCSF website: but without a gov.uk address, and no way to trace who exactly is its source, how would you know it isn’t some kind of elaborate phishing scam?

    Update, 16/12/09: It looks like they’ve now moved it to a .dcsf.gov.uk address, and to a different hosting provider (Every City, by the look of it); which makes me wonder why they jumped the gun?

    Ironically, it was a period consulting at DfES that convinced me it was time to escape the Whitehall machine, and embrace the WordPress community. So it’s great to see them coming on board; and I’m all in favour of departments experimenting with WordPress, whether inside or outside the firewall. There are things I’d certainly have done (very) differently: I wouldn’t have used a directory name ‘wpsample’ for a start, and I’d have tried to fix some of the 112 153 validation errors. It also looks as if they’ve overwritten the ‘default’ WordPress theme, which isn’t wise. And it’s always advisable to use pretty permalinks if you possibly can, rather than number-based query strings. But it’s another step in the right direction, and is therefore to be welcomed.

  • , ,

    Obama's openness directive

    washingtonworks

    The Obama administration’s long-awaited Open Government Directive was published on Tuesday – curiously, in PDF, TXT, DOC and Slideshare, but not HTML? – and seems to have received a warm welcome across the Atlantic.

    What’s quite interesting is its very prescriptive approach. Within a specified number of days, specifically 45, they must have identified and published three new ‘high quality’ datasets in open formats. Then, by day 60 (ie 6 February 2010), each department must have created a new page on its website on open government issues, at a specified URL: http://www.[agency].gov/open. This page must include mechanisms for public feedback and quality assessment; a plan for ‘how it will improve transparency and integrate public participation and collaboration’, to be updated every two years; the annual FOI report; and regular responses to public input. (I’m almost surprised they haven’t offered a wireframe for the page.)

    The document goes on to outline what it expects to see in each departmental openness plan, including of particular interest:

    • proposed changes to internal management and administrative policies to improve participation
    • proposals for new feedback mechanisms, including innovative tools and practices that create new and easier methods for public engagement
    • proposals to use technology platforms to improve collaboration among people within and outside your agency
    • innovative methods, such as prizes and competitions, to obtain ideas from and to increase collaboration with those in the private sector, non-profit, and academic communities; and
    • at least one specific, new transparency, participation, or collaboration initiative that your agency is currently implementing (or that will be implemented before the next update of the Open Government Plan).

    Very specific measures, milestones and expectations – plus mandatory innovation across the board. Would it work here?

    For more, read TechPresident.com’s analysis of the directive, and its implications.

  • , , , ,

    All so familiar in Oslo

    If I take one thing home from my trip to Oslo, it’s the fact that we’re all seeing the same opportunities, and facing the same hurdles. Today’s day-long seminar on innovation in and around government kept coming back to freeing up public data – oh, including maps. Sound familiar, anyone?

    The opening session was Hakon Wium Lie, CTO of Opera (and the guy who invented CSS), making the case for freeing up taxpayer-funded data. As it was in Norwegian, I didn’t get much of it: thankfully, it was essentially the same content covered in this article (complete with Google translation).

    I talked about the UK experience of openness, open source and consultation: lots about WordPress, Commentariat and activism – familiar stuff to readers of this blog. With the Smarter Government paper on Monday, all the social media activity around Copenhagen, and the Tories’ forced commentability on the leaked IT strategy, I wasn’t short of timely examples!

    Then it was Nikki Timmermans from the Netherlands, talking about their education ministry-backed Digital Pioneers fund which has supported 150 projects since 2002. I was particularly taken by one particular project they funded: effectively a ‘dating service’ for their MPs… by which I mean, it tries to match you with MPs who share your interests, based on voting behaviour, beliefs and background, even favourite football team. It’s an interesting challenge to the idea that the MP who best represents you is the one for the patch of your land where you go to bed each night. I wonder if it’s something we could put together in the UK?

    The final presentation was by Olav Anders Ovrebo. It was in Norwegian. I didn’t get much of it. But he works for this website which is WordPress based. So clearly he knows his stuff. 🙂

    After a (really very nice) sandwich lunch, we broke up into small groups for discussion. I couldn’t help feeling guilty, forcing the groups I joined to speak English; but they all coped much better than if I’d been asked to cope in Norwegian, and were too polite to complain.

    Many thanks to Bente Kalsnes of social platform Origo for inviting me to participate. I hope the examples and experience I brought were useful; and I’m certainly taking away plenty of food for thought myself.

  • A Lot Of Orange

    Doing a bit of research for my presentation in Oslo tomorrow (of which more later), I came across a somewhat surprising figure in Hansard.

    Directgov cost us £30.7 million in the year 2008-09 – well over double what it cost us the previous year. Of that £30.7 million, £7.48 million went on ‘advertising, public relations, publicity and marketing’. As one wag pointed out on Twitter, that would buy you a lot of orange.

    Did it work? Well, the best comparison I data I can lay my hands on is Hitwise market share, courtesy of Public Sector Forums – and Directgov went from 9.14% of ‘central government’ traffic in March 2008, to 17.02% a year later. Of course, that doesn’t mean traffic has doubled… and it has to be seen in the light of web rationalisation, whereby Directgov is eating other websites.

    Sadly, the last PQ on Directgov traffic seems to have been in December last year; and DG’s own traffic page only quotes the last 3 months. And even then, the numbers are curious to say the least. They can’t provide a unique user figure for August this year. And somehow, between September and October, unique users more than doubled, whilst visits fell by 8%, and page impressions fell by 11%. Eh?

    dgtraffic