Via the indefatigable David Bath comes word that Lindsay Tanner has taken up blogging in limited form. Right now it’s a bit officially-flavoured, but that may change.

I don’t think that single ‘official’ blogs are the best approach. It is useful for top-level officials to blog in some capacity, but an open approach to blogging works best when any and all officials will be allowed to blog on their area.

Right now there are only a few ‘superblogs’ for the whole of government; naturally this means that they attract interest only from the most dominant topic. For instance are hundreds of comments below Tanner’s first post slamming Conroy’s damn fool ‘clean feed’ notion.

This means that topics that are niche don’t get any sort of look in. Government bureaucrats working on some obscure topic won’t get covered by the centralised blog, and so won’t get to talk to people in their area of work through the blog. They often do so now, but privately, in a way that leaves no open records.

Microsoft and Sun allow their employees to blog – in fact they provide the tools and hosting. Sun has bloggers up to and including the CEO and his executives. People interested in the future direction of file systems get to talk directly to Sun’s file system coders; people who want to discuss Microsoft’s next version of Windows now have an abundance of sources of information of the hows and whys of the project.

A big centralised blog can’t do that. You need to allow individual blogging under the corporate umbrella.

An example turned up on the same blog – an anonymous mandarin using a corporate group name posted this bit about opening up government data.

This leads into a segue on my next concern about government IT / data policy. Any and all data should be as accessible, searchable and queryable as possible. Says the linked post:

At this point, you may pause and wonder-why would the Government want to give private companies a ‘free ride’?

For which I left the comment:

1. The process of opening the data sets for public consumption improves government’s ability to mine its own data.
2. Once the data has been integrated, the cost of making it available for public use is very low. Most of the cost is sunk already.
3. A cost-recovery approach to providing data will create disincentives to produce any data without identifying a ‘market’ first. Yet we can see from the example of ‘mashup’ services that the best ideas come about after, not before, the data is available.
4. Data transparency for government is akin to transparency and accountability generally. It is not an optional extra; it is an essential component for the healthy functioning of our democratic system.

For these reasons any and all possible data should be released, all new data should be indexed in a way that conforms with DIRKS and similar, and all data should be presumed classified as being for-release, rather than needing a lengthy and expensive FoI process to tease out.

I will watch with interest to see how this develops. Though I’m not holding my breath for fast movement. This is the Canberran leviathan we’re talking about.

Elsewhere: Robert Merkel on the same topic.