While I am a big fan of Luke's and agree with his view that transferring government websites to Facebook, granting partial control over them to a foreign-jurisdiction company, is not a good idea, I find it harder to agree with Luke's point on centralising government websites and employing a common look and feel.
I've never been a fan of the 'one site fits all' approach of the UK Government's attempted Directgov website - or a supporter of the view that all government sites should have a common look and feel.
Because websites need to be designed to meet their specific set of goals within the constraints of the needs and preferences of their key audiences.
Where the goals and audiences are different, the websites need to be designed and operate differently.
Even when the goals and audience of two separate websites are similar, there can be good reasons to solve the 'problem' of usability and quick access to key information in different ways.
Web design is an art as well as a science. There's often multiple ways to achieve a good outcome, not one single approach that is best. This means that a government that did lock itself into a single 'right' website look and feel may find itself in a blind alley over time, requiring a huge shift in design to jump onto a more future-proof track.
When I commented on Twitter about my views I was told that a common look and feel made citizens more comfortable that a website was 'official'. This is quite a useful technique in the real world, where standard uniforms are used in a number of government professions to convey officialness and trust (such as police forces).
However online governments cannot trademark a given 'uniform' design for their websites, leaving it open for others to employ a similar or identical layout in order to mislead people into believing they are official websites.
The best safeguards of 'officialness' are those we already use - a common crest (where legal action can be taken to protect it from fraudulent duplication) and the use of a common domain '.gov' which is unavailable to anyone other than government agencies.
These two safeguards ensure that anyone visiting a government website can be assured that it is owned and maintained by the government in a way that a common look and feel cannot.
I always try to keep in mind that government websites are not common places for citizens to visit. Citizens only go to government sites for specific purposes - to find information on a given topic, to access a service or to report an occurrence.
Meanwhile government web staff visit government websites all the time, particularly their own.
I've generally found that while government web teams can point out all the flaws in their sites, visitors (who may go to the site once a year) don't notice them and often have a much more positive view towards government sites than do the internal experts.
I've yet to see evidence that citizens want a single website for government, at any level. What they do want is to find the information or service they are seeking quickly and easily. Google has become the front door into many websites - including government sites - because it meets this need.
Why should government invest a cent into replicating what search engines already do well? We could better invest our money into ensuring that when people get to our sites that the content is current, relevant, written in plain English and fully accessible.
Touching a little further on the concept of a single central government site, often the structure of government works against this approach anyway.
Agencies are funded separately, managed under different laws and often have restrictions on how and when they can share information.
They have widely different needs to engage the public and generally need to control their own web presences in order to maximise their flexibility when the environment changes.
Moving to a single content management system and single website poses a number of challenges for operational management structures, flexibility and funding.
Do all agencies forgo some funding for websites to fund a central agency web unit?
How does an urgent ministerial need (which requires the equivalent of a website today) get fulfilled in a timely manner? How does the central team prioritise development work, and who has access to content - and at what level.
There's just so many questions as yet unconsidered - even in the UK's Directgov model.
While I hate the proliferation of web sites across government, where every policy or program area, government directive and new initiative often 'requires' a new and discreet website, I think we'd be better placed putting a common framework around when and how government websites are built, and developing a central public list of these sites, than attempting to fit all these diverse properties into a single content management solution, central site and common look and feel.
By all means recommend a standard approach (always put the About link at far right, include a Contact, Privacy, Terms and Copyright page, organise content in relation to the audience, not the Department's structure), but don't compel a standard look and feel or central site.
I predict that many agencies would work around a centralised model, simply to meet the government's explicit policy requirements.