Nick Holmes over at Binary Law has posted an interesting review of the new GOV.UK universal website. The UK government's attempt to make web surfing for government information "simpler, clearer, faster."
The website is only partially finished at the moment, with all departments expected to be integrated by April, 2013. Like any large portal, the arrangement of information — along with a dysfunctional site-search functionality, which Holmes touches on a few times — are going to be major hurdles to overcome.
The interesting issue for me, however, is whether they are trying to accomplish too much on a singular domain. Unfortunately, that's a difficult question to answer. Any large portal site is very dependent on its information architecture, and the ability to create a browsable offering. Doing this across multiple ministries, sounds like a lot of conflicting search concepts, and a lot of content to cover. Plus, if Google, as Holmes indicates, is out performing your own in-site search functionality, it sounds like a troubled project from the outset.
On the other hand, if sleuthing government information in Canada is any kind of guideline, then the endless circles of clicking may be a problem for both approaches. It's not uncommon with the cross linking between government websites in this country, to click off of the domain you originally started with. Especially if the ministry in question has multiple domains in play. A well considered site-search functionality, or a quality federated search product for multiple site coverage, would seem in order.
Two facts I don't know (but would like to) are: 1) how many official domains does the government of Canada manage? (thousands?); and 2) Do we have any kind of federated search product covering those websites for our own country?