Unity is on to something with this rant about government Web sites. But I think we can take this a nudge further. Essentially, if you’re visiting a government Web site, you’re doing one of two things; you’re either engaged in some sort of transaction with the government, or else you’re looking for some sort of random lump of information. In the first case, there’s no reason for there to be Web sites, plural; you’ll probably make your way to the thing by a Web search anyway, and once you’re there you want to go through the process as quickly as possible.
The Government is actually trying to get this right by rolling as many transaction functions as possible into Directgov (for individuals) and Businesslink (for organisations). But it doesn’t go anywhere near far enough. Especially for organisations, the ideal type of going through a process involving transactions with the government is to have a machine do it instead. If it’s a transaction – or for that matter, a query of a public database – it should be available as a SPOO (Some Protocol Or Other – an acronym I just invented to cover RSS, XML, SOAP, Web services, REST, JSON, and all their pals) feed/service/whatever, and these should be as much alike as possible across government.
As far as the second type of site goes, you can be certain that people looking for just the right document will be using a search engine; so really all you need is search-friendliness, metadata, and stable URLs so you can find your way back to the stuff. If the stuff is in a database, then there should be some way of querying it automatically.
And finally, there should be a website called developers.gov.uk or something similar with a directory of all the SPOOs, the documentation for them, and a short guide to the legalities involved.
Not that this will discover gunpowder, but it’s a nice timeboxed quick-impact project that a small political party with an interest in freedom of information could do worse than to push, and that would make life mildly less annoying.