Home › betagovuk

Precious Snowflakes

Sat, 2012-03-10 11:56 — Jeni

Disclaimer: As usual, this post contains my personal opinion and does not reflect that of any organisation with which you might associate me.

The other day, I had a lovely conversation with some folks from the BBC about some of their future plans. In the course of the conversation, Michael Smethurst spoke about his frustration when dealing with people involved with particular programmes at the BBC, where every single one of them thinks their programme is a “precious snowflake”, completely unique, that simply can’t be treated in the same way as all the other programmes described on the site.

Michael’s point, of course, is that TV programmes have a hell of a lot of similarities with each other. They all have episodes and cast members and may have trailers or be available on iPlayer. When the BBC models them in the same way, they gain enormous efficiencies in their ability to store and access information about programmes: they can reuse code, share content between programmes, and perform analyses over the aggregated data set. It’s great for users too: they get the same fantastic user experience no matter which programme they are viewing information about, and can apply the experience they gain when navigating pages about one programme when they need to find information about another.

The ability to classify and categorise, to bring order to what seems like chaos, to create a model of the world, is one of the things that marks humans from animals. We can look at a hundred people, with different colour hair and skin; different height and build; smiling, talking, crying, and still call them all Person because the essential characteristics that govern how we interact with them are the same.

But if there’s one thing that the last five long, hard years working with legislation has taught me, it’s that in any vaguely interesting domain, this search for order will always fall down in the face of reality.

Surely, I thought in my naive early days, every piece of legislation is uniquely identified through its type, calendar year, and number? Not so! There are six items for which this is not the case, because prior to 1963 legislation was numbered based on the year of reign of the monarch rather than the calendar year.

Surely the year that is used to number legislation is dependent on the date it is made and written into law? Not so! Sometimes departments forget to register legislation they make until the following year, so it is numbered the year after it’s made.

Surely an item of legislation can only make changes to legislation from the day it is written into law? Not so! There is, rarely, legislation that rewrites history: that says other legislation should always have had different content to that which was originally written.

It has come to the point where I never (hah!) make any statements about legislation of the form “X never happens” or “Y is always true” because there is always, always, an exception.

What this has taught me, as a developer, is the power and necessity of escape hatches. For example, templating languages that provide a method of escaping to code are so much more valuable than those that do not. Similarly, I favour strongly, in the technologies that I use, the ability to extend a common data structure, be it through data-* attributes in HTML, through generic elements such as <span> and <div> or through the essentially open-ended nature of RDF as a data model.

It has also given me a very different view of the world to Michael. Because when you accept that there are always exceptions, you do not see snowflakes as merely crystals of water, but as exceptional, beautiful and, yes, immensely precious.

And this is why I love the web. The web does not force every site to have the same structure or the same look and feel. It does not insist on consistency; it has space for every quirk. And it proves beyond all doubt that it is possible for all these precious snowflakes to exist in a single, global, interlinked information system in which people manage to find not only the information that they need, but also community and connection with each other.

Inside Government

So it is with these eyes that I look at the new Inside Government pages on the gov.uk site and am frankly horrified. Because we’re not just talking about a BBC programmes here, but about powerful institutions, many of them decades if not centuries old, that lie at the very heart of government and how our nation is run. And each of them is relegated to a subfolder of a subfolder of a subfolder, their unique histories and approaches and goals expressed through three pictures on a carousel.

It feels like some kind of Orwellian nightmare: the relentless focus on user needs leading to a future of identikit pages, with no individuality, no character, no clue that behind these pages — which, remember, under the Single Government Domain policy becomes the single authoritative view, the site that represents the department on the web — is a living and breathing institution that manages hugely important parts of our lives. A future in which what each department says and the way that it says it is governed through the Government Digital Service (GDS), in Cabinet Office, the hand of the prime minister. And now we’re talking about local government too?

Let us just look at one example. Last September, William Hague gave a speech in which he described the hollowing out of the Foreign and Commonwealth Office (FCO) by the previous government, a process that scrapped its language school, closed embassies and destroyed its library. He said:

Strong institutions are necessary in civil society, to encourage participation and keep in check an overmighty State; they are necessary to our judiciary and Parliament so that the law is upheld and the making of it respected; but they are also necessary within the State, a point tragically overlooked by those Prime Ministers who have created and abolished departments on a fancy or a whim, destroying as they did so the pride and continuity of thousands of public servants while rendering government incomprehensible to the average citizen. The whole country should know what the Foreign and Commonwealth Office is and what it does, and all those interested in foreign policy at home or abroad should see it as a centre of excellence with which they aspire to be associated.

For most UK citizens, the only point of access to the Foreign and Commonwealth Office is its website: they will not visit King Charles Street, nor any of the UK’s embassies. The department’s web presence is the only way that it makes itself, and its unique role, comprehensible to the average citizen, the only method of letting the whole country know what the FCO is and what it does. And they have content that is completely unique to them: a database of Treaties, a hugely rich set of information on travel and living abroad and a wealth of historical information about the Foreign Office. This simply doesn’t fit in a model of a department as a set of Ministers, Policies, Publications and so on. And if it doesn’t fit, will it simply be excluded, lost from its website like its language school, its embassies, its library?

I could have picked any government department here — each one has its unique characteristics and content — but Hague articulates the case around FCO so well. His message is not the expression of a simple conservative impulse to resist change and preserve the status quo, but about maintaining the integrity of an institution’s identity and independence to encourage participation and keep in check an overmighty State. If we believe in Open Government, Open Democracy and the power of the web to enhance civic engagement then we must, surely, enable each of these institutions to have their own independent voice on the web.

I am reminded of the XKCD:

The two sides of this Venn diagram illustrate two approaches to building a website for an organisation. On the left is the website as an expression of the identity of the institution, on the right the website as a means of satisfying the reason the user originally visited the site. My argument is not that the right side of this diagram is unimportant — in fact I believe it is absolutely essential — but that an institution’s website must cover the entire space: it must provide a mechanism for self-expression as well as catering for its user’s requirements. To enhance civic engagement, we do not need to simply answer the query that led the user to the site, but to encourage and lead them on to see more about the institution that has provided the answer.

It is only the institution itself that knows the self it wants to express, and because the real world is complex and organisations are unique, that self will not fit into any model that we devise. News, Policies, Consultations — of course these are all important to all departments, but they are the tip of an iceberg. Look at the space that FCO devotes to its history on its website: this shouts to the world the kind of reliable, solid and flexible organisation that they are and want to remain. Compare how DECC devotes space to statistics, emphasising its adherence to transparency and evidence-based policy. Self-expression is so much more than changing logos or backgrounds, more than having different content on an About page, it is about making space for the things that are important to you.

“But but but!” I know the arguments. We must cut costs, stop the uncontrolled proliferation of government websites; we must improve the quality of the government’s presence on the web, present a unified view, make it easy for users to locate content without knowing where to look. The vision we see expressed through Inside Government is but the natural conclusion, the end of that slippery slope. But it is the great slippery slope fallacy that everything must be taken to its natural conclusion, that because 750 websites is too many, one is enough.

Possibly the biggest irony of the gov.uk beta is that while it is delivering a Single Government Domain — everything is to be found under www.gov.uk — it does not seem to address the core reason stated for providing it. In Martha Lane-Fox’s letter to Francis Maude, which kicked off this whole endeavour, she said:

Government publishes millions of pages on the Web, via hundreds of different websites. Most of these sites are still run as silos within departments. This fragmentation leads to significant duplication of functions and technology, and means the overall user experience is highly inconsistent.

Try searching for ‘Single Government Domain’ on the main gov.uk site or for driving test centres on Inside Government. The searches do not work (though Inside Government does give you a link that enables you to search the other silo). The result pages are completely different in feel except for the top and bottom banners. The page on Arrest and Imprisonment Abroad mentions but does not link to the Foreign and Commonwealth Office’s page. Yes, yes, I know that it’s still beta, but these things lie at the heart of the stated rationale for a Single Government Domain: is this the extent of the consistency and integration that we are aiming for?

Yes, it is, because the Single Government Domain policy was never truly about either of these things. Read Martha Lane-Fox’s letter again carefully (my emphasis):

No1O feel it is preferable to go from 750 top level website domains (eg www.cabinetoffice.gov.uk) to a single top level website domain for all of central government.

The Single Government Domain policy, indeed GDS itself, is about control. It is “we will do it for you”, not “we will help you do it”. It is about managing the output of institutions that might keep in check an overmighty State. It is anti-web and it is anti-democracy and I cannot remain quiet about it any longer.

To my friends at GDS: I respect and admire you all. You are incredibly talented and able to do amazing things. You have behind you a level of financial and political support the like of which most civil servants will never see. I know you have joined GDS not just to do work that you love but to do good for the country. This is my plea to you: find a way to avoid this vision. Nurture the exceptions. Give institutions their voice. Treat them as precious snowflakes.

Comments

Tue, 2012-04-24 11:35 — Ben Wheeler (not verified)

Re: Precious Snowflakes

Very interesting article, thanks.

Some years ago, I worked on a long-term project which was funded by the then Public Record Office. When that institution was rebranded as the National Archives, complete with shiny new website, they decided that our hitherto independently-styled and -managed website must be rebranded to mimic theirs in look and feel.

This was far from easy, partly because their design had a horrendously messy implementation, and partly because (of course) it had been designed without any reference to us or how our data delivery might fit into it. It was imposed on us as a fait accompli, and we had to - somehow - squeeze our square peg into their round hole.

We spent a full year smashing our clean, lightweight design into pieces and gluing it back together in order to fit their restrictive, bloated one. I didn’t much enjoy doing it (can you tell?), but I like to think we did a good job.

Possibly too good. What we found when it went live is that users got confused: our site, now a subdomain of theirs, looked and felt so similar to the main site that users expected it to work in exactly the same way, but this was ultimately impossible as ours had a fundamentally different set of functions than theirs. Those areas where we overlapped had been made to work identically, but this just led to confusion where the functionality diverged.

One size does not fit all. And the more distance there is between those responsible for the design and management of a site, and those producing the content for it, the more likely it is that some of that content will be presented poorly, or not at all.

I don’t think people really want all government websites to look the same, or to be in the same domain (URLs are irrelevant to many non-technical users, who nowadays routinely rely on search engines - even to find sites that they visit every day, as evidenced the “Facebook Login” debacle). I think what they want is for information to be easy to find and easy to access. The best way to ensure that is to keep the designers and managers of the website as close as possible to the people producing the information. By all means have standards to ensure best practice, but keep them as minimal as possible, with a mechanism for those bound by them to suggest changes if they find them too restrictive.

And let different things look different, because that helps people to realise that they are different.

Wed, 2012-03-21 15:58 — Paul Clarke (not verified)

Re: Precious Snowflakes

It's not about the websites.

It was never about the websites.

The whole “convergence”agenda—or “bonfire of the x,000 .gov.uk websites”—has always been a cipher. A cipher for the (usually external) moderniser's wish to “just make it all simpler”. Who—as an outsider, a consultant, a technologist or in opposition—doesn't see the opportunity to do things better?

And how often in the course of history does a technology come along that's disruptive enough to allow you a shot at really simplifying things? Permanent, radical change. To start from scratch. To throw away the flim-flam, the skeuomorphs, the layers of bureaucracy and the artificial boundaries and just do it, well, more simply?

The analyst's dream: government reduced to its atoms of content, and its deterministic processes. No faff: cut straight to the information, the call to action, or the interaction. Nothing else. Beautiful, unfettered technocracy. A place for everything, and everything handled once.

Putting services online, creating new digital services, recreating organisations as online entities—approach or describe it how you will—this has always been seen by influential policymakers as a magic bullet to “transformation”. If, the argument goes, the future of information and interaction will be online, control that experience and you control the soul of really big, important things like public administration and democracy. Then it doesn't matter if you have a host of legacy functions sitting behind a polished, unified front-end. Out of the sunlight, they'll wither. They'll have to. They won't be able to justify separate, divided existence. Radical reformation will be assured.

It rarely works like that, of course. For a whole host of reasons: not least that a reductive approach to an organisation as the sum of its interactions misses many of the nuances of organisational design that are about human relationships and status, not information flow. Whitehall departments weren't oak panelled because that was the most effective or cheapest way to decorate the walls. You still have to have models of governance and accountability, no matter how intense your aversion to bureaucracy. Who governs the governors?...and all that. And, more prosaically, we've still never really squared away the rational link between “having your own website” and “legitimately/permanently existing as an organisation”. Not in hearts and minds, anyway.

So be aware that "single domain" and other talk of simplification is but the veneer on a far more fundamental project. Be honest about that, and you might just have a chance of making such change happen.

Thu, 2012-03-15 14:46 — Ross Ferguson (not verified)

Re: Precious Snowflakes

I passionately agree with your point that departments have a unique ‘living, breathing’ identity and must be able to use it online.

But these days (and in the days that I can see up ahead), less and less of this ‘personality’ is dependent on departments’ corporate websites; instead it manifests and develops through their owned and earned presences in the social web. Not that departments’ sites should become dry and officious, but it is easier and more meaningful to engage where the people are in blogs, social networks, hubs and other social forums around the web.

I say that based on my experience of researching different settings for web-based engagement between government and the public. As well as firsthand experience of delivering engagement projects for government departments, which has included examples both on- and off-platform.

I’m the current technical director of the FCO’s web platform. And we are one of the 10 departments probing and critiquing the beta as participants.

We are pretty pleased with the current FCO.gov.uk. We’ve worked hard for over 10 years to make sure the site meets user needs and delivers on the department’s obligations and objectives. It’s heartening and humbling to get the feedback and see the usage increase year on year.

We are also critical of it. And it is because we want to have a more effective and efficient digital presence that the transformative impact of GDS and Gov.uk is welcome. The arrangement whereby GDS leads of the platform and the user experience, and the departments are freed up to own and publish their content and engage with users is probably long overdue.

The beta of ‘Inside government’ doesn’t give rise to fears for the future of the treaties database or the travel advice or the ability to consume our content in 40+ languages. From practical experience of the beta, I’ve seen GDS get as passionate about departments’ user needs, functional requirements and content as the proprietors themselves. They have also helped departments take a long hard look at what they are doing, explore the reasons for it and what other ways there might be for delivering ‘simpler, clearer and faster’.

I am reassured that GDS will work in partnership with the departments to amplify what they do rather than muffle or obscure it. Even if that was what GDS was deliberately or accidentally doing, the departments are savvy and authoritative enough to prevent it. I am so convinced (from an initial position of scepticism) that when my secondment at FCO ends, I will be joining GDS. But when I get there I will not be hanging up my departmental hat.

Fri, 2012-03-16 10:41 — BarryT

Re: Precious Snowflakes

I do agree about the value of social networking as a means of engagement. BUT my alarm rises nevertheless. I happily use twitter, but I DON’T use facebook on principles of walled garden and privacy concerns. So, especially as part of “my” government, please don’t put any important content (which I should see) anywhere other than in the open spaces of the web, open to all to view. And, incidentally, as open data, please, not proprietary forms and formats.

Thu, 2012-03-15 08:03 — Pierre Col - Antidot (not verified)

Re: Precious Snowflakes

I agree with you: even if each TV program, newspaper article, blog post has its unique content and value, the ability to link all kind of content through semantic web aproach, and to store those massively linked metadata in massively linked RDF triple stores creates a new opportunity for the humanity. Semantic Web brings us the concepts and tools to really leverage all the content (data, information, thoughts) produced everywhere in the world by creating new knowledge objects and new informations that will help us understanding our world. And those new information objets are even more precious that the “precious snowflakes” they are aggregated from.

That’s why wh need to take this semantic web opportunity!

Tue, 2012-03-13 11:27 — Carl Haggerty (not verified)

Re: Precious Snowflakes

This is a very interesting post…i won’t comment on the whole, but merely provide context in relation to the local government angle you refer to.

As author of the blog post on the GDS blog about a Local Gov GDS, I think there is a misunderstanding in that people are assuming (wrongly) that my suggestion for a local GDS is that we create a central “authority” to tell us all what to do and that authority is GDS itself… That could happen anyway without local government involvement depending on national politics of course.

My point is really that the last 10 years of external influence and best practice via people like Socitm (who provide useful services) have failed to actually improve the quality of the local government web estate.

The issue for me is that local gov web professionals as a community are very talented and committed people but we ()i include myself in that community too) are all misguided, we follow star ratings when we should be focusing on local peoples needs and creating a “local” site for “local” people…

These kinds of things are often undone when a “maverick” councils creates something different and changes the game and raises the bar…but not always in the right direction.

However there are some broad approaches to managing a local gov web estate that across the sector we don’t have any consistent views or even “loose” standards around for example domain name proliferation, standard URL’s for common services.

My suggestion is that GDS as a body of people can help and advise local government web professionals on some aspects of building better websites and that requires engagement, which doesn’t appear to be happening yet (this isn’t a problem right now) if we as a local gov community wanted to create an equivalent to GDS what would it do and how could we do it? This was the kind of discussion i wanted to generate.

Perhaps my post was poorly written and I failed to get my point across - what it did do was stimulate and spark debate and discussion…

Tue, 2012-03-13 01:24 — Steve (not verified)

Re: Precious Snowflakes

Thanks for opening up this debate – I have to say I don’t agree with your reasoning, but I think it’s a really useful point of view for GDS to hear.

Like you, I think the FCO is a fascinating institution. But I suspect we’re in the minority here. And if I want information on its culture and history, I can get it from books and TV documentaries. Just like if I want to know how well it’s delivering, I can pick up a newspaper, check out an independent blog etc. You can’t really expect any institution to offer a genuine insight into itself and its workings – the incentives all point in the other direction.

So how much would we really lose by adopting a more uniform approach? For my money, it’s actually a great opportunity for government to become more selective about what it publishes. After all, if we’re spending public money we should focus on providing information that’s wanted/needed by the public, and can’t be got elsewhere.

But I do think there are considerable challenges ahead for GDS, even with the clout at its disposal: it’s always going to be a powerful temptation for government to publish content for tactical reasons rather than intrinsic merit.

All that said, I think you’re onto something when you question the assumption that “because 750 websites is too many, one is enough”. If a separate url is a better solution for the user and/or significantly easier for GDS to deliver, why not be pragmatic? It doesn’t stop you realising cost savings by re-using templates, infrastructure etc.

And it means you can apply your time, energy and developer/editorial talent where it’s needed most – helping to define what we do and don’t want from government websites, and making sure that what’s done is done well.

Mon, 2012-03-12 20:26 — Gannon (J.) Dick (not verified)

Re: Precious Snowflakes

Sorry to go all scientific on you Jenni, but Venn Diagrams have a hidden reasoning trap. If you ever had Organic Chemistry, and/or have the scars to prove it, you’d recognize any two ring Venn Diagram as having two isomeric forms … take one ring and twist 90 degrees clockwise or counter clockwise and you’ll see what I mean. The two forms are mirror images. My point is simply that the two approaches to building a web site have different results if “approached” from the right or the left. This is how Nature sorts it out: Mashups and Mashups (tech.).

Domains (as repositories) use a communications language with the Public and a content language to describe the behavior of a Person in their own words. The IETF RFC5646 misses this distinction and says “shortest code” as if the Venn Diagram was in free rotation. It is not clear that “en” and “eng” are two different languages. However, for example in the US, 75% of the people speak English at home and the other 25% speak 110+ different languages. For Social and Public Health Policy, both language and local “Rule of Law” need the same data set coverage.

With work, you can tame the propensity of Bureaucrats to use their own language (Codes) to control what People say and what the Public hears.

eGov Portals

Mon, 2012-03-12 20:09 — Sean McGrath (not verified)

Re: Precious Snowflakes

Jeni,

Paul Downey (Hi Paul!) has already mentioned Conway’s law. It runs deep in Government all around the world. How to find the balance between snowflake and Model T Ford….It is possible to have cake and eat it to but requires a lot of work at the data model layer to make it happen. Homogeneity in the data model for machines but allowing a thousand-flowers to bloom at the publishing and service-deliver ends i.e. front-ends. We have the pieces we need these days to make that happen: REST, XML/JSON, JQuery driven Portlets. etc. The snowflake requirements need to be an intrinsic part of the design. If not, the organizations over time will just subvert the strictures of the “wired hierarchy” and go their own way again.

Moreover, Government’s institutional structure is not static and therefore challenging - at best - to decompose hierarchically. Departments come and go, get merged, split up, have their names changed etc. Any taxonomic decomposition of the overall Organizations namespace needs to be capable of dealing with that or it will atrophy over time.

This means permanent identifiers. A lot of thought needs to be given to these and I would argue that pressures to organize them into a multi-leveled hierarchy needs to be resisted :-) Otherwise, the data modeler is playing their part in proving Conway’s Law.

Case in point: In KLISS, there are permanent identifiers for legislative committees. E.g. House Appropriations : http://kslegislature.org/li/b201112/committees/cttehapprprtns1/

At the end of the URI, see the “cttehapprprtns1” part? That is a permanent, unique identifier out of a flat namespace of committee identifiers for kslegislature.org. We try to make the identifiers reasonably short yet reasonably understandable too. The URI has a whole is crafted to be permanent. The b201112 part uniquely identifiers the biennium for the Legislature. This time next year the link will continue to work but it will point to the quiescent record of the 2011-12 biennium materials.

Imagine next year that this committee changes its name or merges with another committee or whatever. If that were to happen, we would create a new identifier in the flat namespace. Any thematic organization of the names into some sort of order is left to the “many flowers blooming” side of things. Morever, we never remove old names. On the web, “old” URIs constitute debt that needs to be constantly serviced with resource representations. Cool URIs don’t change.

Some might like to see committees associated with Chamber. Others might like to see them by thematic area, or departments involved, or chronologically by meeting dates etc. All “views” are considered equally valid with no “tyranny of the dominant decomposition” embedded in the naming convention.

Hierarchical arrangements of agencies/departments come and go in Government. Titles and abbreviations change meaning over time. Most Government identifiers are contextual in that most challenging of dimensions : TIME. Who is the Minister for Railroads? Where is the department of Post and Telegraphs? Who is the Member for Tyne and Wear? etc. etc. Like human language, organizational structure is a living, breathing, ever-changing thing.

Regards, Sean

Mon, 2012-03-12 18:16 — Phil Archer (not verified)

Re: Precious Snowflakes

Thank you Jeni, this echoes a couple of things I’ve been thinking although in very different spheres and on a much less grand scale.

We were in the room together when one of them came up not so long ago - making dates fit xsd:datetime. If you have data about an event “in August 1984” that’s accurate of not very precise. Changing it to 2004-08-01T00:00:00Z makes it a work of fiction but hey, it matches the reg ex now so that’s good, right? No.

In an entirely different field, there is a continual cry among some people to come up with a sort of pan-European, if not global, classification system for films, TV programmes and games. Having worked on one such system for many years I know first hand how pointless such a move is. How can you possibly improve upon the Irish Film Classification Office’s description of Borat: “Politically incorrect humour on a seismic scale” by mapping it to some universal yardstick that measures the presence/absence/frequency of bad language, nudity etc.?

But… where I hope gov.uk can help is in presenting data that can be systematised in a systematic way. The current list of Home, About, News & Speeches etc. looks good to me and one can imagine that flipping between the departments is made easier by that regularity - but IMO that should be a feed supplied to gov.uk from the departmental sites that, as you say, retain the richness and appropriate presentation of themselves. Maybe I’m greedy - I want the homogenised ice cream and the precious snowflakes.

Mon, 2012-03-12 17:11 — Bernadette Hyland (not verified)

Re: Precious Snowflakes

Thank you Jeni. You remain one of the leading lights on how to effectively and thoughtfully publish Government content. I’ve forwarded a link to this blogpost to customers, eGov and US data.gov lists hoping we can avoid similar pitfalls. Creating useful government sites will take a lot of iteration. If a government team building the ‘one ring to rule them all’ thinks they can do it alone, they’ll wither in a small circle with their hubris. Let’s all remember what made the World Wide Web the great equalizer — ease of contribution and access. It is through leveraging the strengths of the developer community, the open source community, researchers & commercial firms that we’ll achieve a sensible and useful site with government content, one more like the right hand side of the Venn diagram.

Thanks for sharing your wisdom & insights so we don’t make the same mistakes rather, build and extend which is what the Web (of Data) is all about. Please keep up your valuable work & musings.

Mon, 2012-03-12 17:05 — Mick Phythian (not verified)

Re: Precious Snowflakes

Jeni, A great thought piece! As part of my academic work I came to the conclusion that much of egovernment was flawed for the simple reason that ‘government’ was far from being as simple as politicians believed it to be. There was, to employ Giddens’ phrase ‘structuration’. The reason why we have problems joining health and social care in the UK was the way politicians divided up the spoils in the creation of the NHS. Prior to that there were the effects of the Poor Law Code and everything prior!

When politicians come up with these solutions they see the smallest surface of all this and think anything is possible but extracting things out of their historical, social and other contexts will and does rebound.

I’m going to enjoy re-reading this more than once!

Mick http://greatemancipator.com

Sun, 2012-03-11 23:44 — Paul Downey (not verified)

Re: Precious Snowflakes

I don't know Jeni, there's a lot in this post I don't agree with, or recognise, even after a few short weeks working in GDS. The only thing I can really say at this point is I've been lucky to witness first hand how the "Inside Government" beta is being developed, and that's in very close collaboration with people of multiple disciplines from across government. There's a general sense of excitement from the departmental staff I've met arising from the opportunity to work directly alongside Web designers, experience and front-end developers and copywriters to build the site together; a very new and more flexible way of working compared to old procurement and publishing procedures, as one departmental "Policy Wonk" told me.

To answer your primary concern: I think there is general recognition that the publishing platform and processes must work well for both citizens and individual departments or Conway's Law will indeed out.

Mon, 2012-03-12 10:06 — Jeni

Re: Precious Snowflakes

Hi there Paul,

Honestly, I don’t doubt any of how you describe the workings behind the scenes on Inside Government: all the GDSers I know are amazing, good people (not just good at their jobs, but good in their hearts). I am sure that people within departments, whose own ability to do anything with their websites is stymied through cuts in their budgets as well as systems and procurement horrors, are over the moon to have the chance to work with experts such as yourselves.

My fears aren’t really about how GDS is working, not at the moment. They are based on the fundamental policy direction that sees a department’s address on the web going from:

http://www.fco.gov.uk/

to:

https://www.gov.uk/government/organisations/foreign-and-commonwealth-office

and what that says about the importance of the department as an institution, the locus of control in the management of a department’s website, and the way in which evolution and innovation occurs within the government web space. My concern is that the current great centralising impulse in the name of user experience undermines the subtler and deeper benefits that the web brings through its distributed nature. And I am worried, in the long long term, about the working relationship between CO/GDS and the various organisations who will completely rely on it for their web presence.

I think the good people who are at GDS now have the chance to embed technical architectures and processes which could mitigate the worst of the consequences of this policy in the long term; I’d just really like to see some evidence of that thinking happening.

Let’s discuss it over drinks some time soon :)

Jeni

Mon, 2012-03-12 19:42 — floehopper (not verified)

Re: Precious Snowflakes

http://www.gov.uk/fco also takes you to the FCO page on the new site.

Sun, 2012-03-11 22:25 — Jim Fuller (not verified)

Re: Precious Snowflakes

I think you have struck a universally important chord that goes far beyond the disruption occurring in governments.

Striving for homogeneity can realize efficiencies but at the cost of precision. We could just as easily say we should all speak a ‘lingua franca’ which would result in magnitudes of benefit, but what do we lose in translation ?

Being able to embrace differences is at the root of so many issues … its a hard challenge and all too often its easy to react with fear (and subsequent control) when faced with it.

I would rather embrace all the differences, promote a ‘long tail’ of understanding and eschew with a ‘winner takes all’ approach. The opposite approach never works anyways.

When I hear about 100 million (insert currency) being spent on something like human beings communicating to each other I think to myself this is a good thing, we should spend more on that. In fact, governments have no problem spending much vaster sums on things shrouded in secrecy with debatable impact on our lives.

The world (and industries and governments) are currently being disrupted by the interwebs and I am glad that smart folks in the UK civil service and people like Jeni are passionate about changing governments for the good … its not often you get enlightened (excuse the term) and high calibre people willing to pitch in and voice an opinion where it would be easy to say nothing.

The less power government institutions have and the more power vested to the people the better. The less government and the more tools with which the ‘governed’ can get on with it by themselves is how to reduce costs.

Sun, 2012-03-11 19:38 — Adam Retter (not verified)

Re: Precious Snowflakes

Jeni, an excellent piece and well argued. I think you have many valid points here. In your comment to Tom, you said -

“The way GDS should drive down costs and increase quality is by providing them with an integrated range of shared services, platforms and tooling from which they can select whatever level of technology they need.”

Honestly, I could not agree more! At the same time, I am smiling, as that is exactly what we tried to do between 12 authorities in Devon in ~2004 as part of the Governments e-Government initiative. Unfortunately whilst we had some successes, ultimately we did not have the political clout to get the long-term buy-in required.

Sat, 2012-03-10 19:31 — Phil Wilson (not verified)

Re: Precious Snowflakes

So, I do work at a university, and I can tell you that that XKCD is pinned to the wall of just about everyone on campus who works on the web :)

The way I view that Venn diagram is about priorities - you must be able to deliver the stuff on the right successfully in order to be able to deliver the stuff on the left, since whilst the left side has value, it’s mostly expressed through the context of the right.

Sat, 2012-03-10 15:00 — Michael Saunby (not verified)

Re: Precious Snowflakes

Brilliant.

But….

Should such projects be actively halted, or allowed to wander and find their own place in the world?

If it were me then all the stuff on the left side of the XKCD venn diagram belongs on facebook and the right side as APIs that others can present to users, robots, aliens…

Sat, 2012-03-10 13:23 — Tom Loosemore (not verified)

Re: Precious Snowflakes

Thanks for a very thoughtful contribution, as ever. But while I agree with some (most) of your words caution, some of your key assumptions are wrong.

Firstly, I couldn't agree more about the search for complete order falling down in the face of reality. TheyWorkForYou.com taught me about the messiness of the Parliamentary record, while assorted programme-metadata work at BBC & C4 taught me a similar lesson. Some messiness is good (it allows snowflakes to be unique), but a *total* mess is bad (you couldn't find out how your MP had voted before TheyWorkForYou).

We should always seek an optimum balance between structure and messiness. I tend to err on the side of "just enough structure", rather than "just enough messiness".

The trick is, as you suggest, to try to find the balance.

Hence the objective of the INSIDE GOVERNMENT phase of the beta was to deliver a *private* beta of a platform to meet 80% of the *shared* departmental publishing needs. Shared departmental needs are but a subset of total departmental needs, albeit a more than large enough subset to justify a shared platform. By way of analogy, note that all BBC programme websites are now run off a shared core platform, with *huge* variety on top. That's our model.

I will admit that the frantic tenor of your post made me wonder for a moment as to whether we'd done the right thing in making this phase of the beta public, albeit only for 6 weeks. I'm still glad we have, mainly because the push to change mode and go public with this phase came from our partners in other departments, who wanted to test how such a platform would work operationally. We've not even tried to join up the two halves of the beta yet, and there's *tons* of publishing stuff left to do that we've barely begun to explore with departments.

And I'm afraid you've just plain wrong about GOV.UK being "about managing the output of institutions that might keep in check an overmighty State."

Departments will still edit their departmental websites. GDS will not, neither will No10. You've jumped to an entirely wrong conclusion, here.

GDS is about doing stuff "for you, with you and by you". As the list of people from across departments attests, this one is "with you".

Also, Departments *are* the State. They can't, by definition, hold themselves to account. On the other hand, making it easy for users to see all departmental policies, consultations, news, ministers, speeches etc, as the Inside Government beta allows, is vast improvement in accountability.

Our aim is to create a GOV.UK that makes life simpler, clearer and faster for users, and cheaper & more innovative for government.

We will nurture the exceptions, but not at the cost of the level of confusion currently suffered by users. Or the £100m/year this chaos currently costs each and every one of us.

Sat, 2012-03-10 22:47 — Jeni

Re: Precious Snowflakes

Thank you for your response, Tom. As ever, I hope and try to be an honest friend. You released the Inside Government beta to the public for feedback, and this is mine. I do understand that these are early days, but I think it’s important to raise these concerns now because it is now that the direction of these sites will be shaped, the underlying assumptions put in place that will determine the future state of the government web. I am not frantic, but I am concerned that in a year it may be too late and I do not want to be wishing that I had said something sooner.

You are absolutely right that I do not know or understand all the plans that you are forming within GDS. I can only judge based on what I do see, based on the sites and words that are made public.

When you say ‘departments will still edit their department websites’, I take it you are referring to the set of pages that is available within gov.uk about the department. This aspect of the Inside Government beta is an accurate representation of the place that departments will have within the government web space under the Single Government Domain vision, correct? Martha Lane-Fox’s letter says:

Ultimately, departments should stop publishing to their own websites, and instead produce only content commissioned by this central commissioning team. There is no need for a major migration of content from existing departmental websites, they should simply be archived or mothballed when essential content has been commissioned and included in the new site.

So it is not so much that departments have their own websites, rather that they provide content to a site that is managed by GDS and by Cabinet Office.

I know that this is a subtle distinction but to me it is an immensely important one. It is the difference between being able to edit parts of your Facebook page and owning your own domain.

Departments are parts of the State, yes, and they hold each other to account. They have different aims and objectives, and power constantly shifts between them. A view of government that does not give space for these differences may be less confusing to users but it is hardly accurate or transparent or accountable. What’s more, as the Single Government Domain spreads, isn’t its aim to take in arms-length bodies? Organisations such as the Audit Commission and the Environment Agency have clear remits to hold the rest of government to account.

Concretely, I think government organisations should be in a position where they have complete control over their own websites and their own domains because those are their identity on the web. The way GDS should drive down costs and increase quality is by providing them with an integrated range of shared services, platforms and tooling from which they can select whatever level of technology they need. They should be helped with assessing and meeting their user’s requirements, helped with improving the quality of their information architecture and user experience, helped with procurement and licensing and all those other things that have bogged down their ability to improve and innovate.

For sure there are huge technical and organisational challenges there, but don’t tell me your team can’t meet them.

As you say, the choices here are about balancing priorities. My argument is that shifting the balance so far towards the overt user requirements for department websites (often specialist, and not of immediate relevance to the vast majority of consumers) comes at the cost of losing something essential in our democracy. No matter how important it is to drive down costs and avoid confusing users, that feels like the wrong balance to me.

You have put together an amazing team, and you have captured an astonishing amount of political capital. From what I can see, innovation within the government web space and outside GDS has more or less ceased while everyone waits to see what GDS will do, and my impression is that people are in some way scared to question or discuss the GDS vision, in case this wonderful and exciting bubble of proper investment in and rethinking of government IT suddenly bursts. Certainly that has kept me from writing about my doubts about the fundamental direction of the Single Government Domain and GDS before.

But GDS doesn’t need yes men; none of us do. We need friends to challenge our assumptions, point out our blind spots and make us think about what we’re doing. If my concerns are misplaced because of my lack of understanding or assumptions that I’m making, then great, I am pleased to hear it, but we do all need to be having this conversation rather than suppressing it.

Sun, 2012-03-11 13:47 — Jeremy Gould (not verified)

Re: Precious Snowflakes

Goodness. What an excellent piece. Don’t agree with all of it but we need much more of this constructive dialogue and criticism to drive the agenda forward in the right direction. Glad somebody still has fire in their belly.