So what about that Public Data Corporation?
I think it should be opposed with full force. Why? Well, the fact they want to structure it as a "corporation" is suspicious in itself - it's going to be tempting to declare it a trading fund and force it to make profits, even if only in the silly Tory sense of forcing other bits of the government to play at shops and pass bits of their budget to other budgeting entities. This would send us right back to 1996 - no access unless you're rich, thanks. The model here is the Forensic Science Service; this was trading-fundised, so that money from the policing budget was sucked back to the Treasury with every DNA test, and has now been sold to the private sector. (Expect to need a DNA-driven innocence project within years two, I would.)
Further, the point of the exercise isn't defined at all well. Sometimes, you might think this was an opportunity. If something is ill defined, there may be the opportunity to define it. However, it is hardly a controversial statement that the current government has something of a record of introducing ill-defined and poorly thought out wheezes that end up in a horrible mess. It's therefore prudent to assume that this particular wizard wheeze is just as poorly thought out as all the others, and will end up in a horrible mess unless someone intercepts it. Just look at what they've done to some perfectly good projects. Hopefully, that's "intercept" as in the way the air force do it - splash it in the North Sea.
Also, the current position makes sense; data.gov.uk lives in the Cabinet Office. The Cabinet Office's core responsibilities are for the machinery of government and the Civil Service. Data disclosure and privacy issues are fundamentally administrative. The dossier is where it should be, and conveniently close to seriously influential people. Bumping it to some sort of thingy-wotsname quango based hundreds of miles from Whitehall would be a defeat.
Finally, the best reason to oppose it is that if we spend the next three months refining our response and debating the philosophical nature of privacy, whatever it is they want to do with it, they'll just do it and the damage will be done. If it turns out to be a reasonable idea, it's always possible to welcome it later. But the flip side of "working with" them to improve it is opposition. It's the early no that buys later influence.
Say no to privatisation and the back to '96 agenda. Say no to the Public Data Corporation. And if you're working with government datasets, scrape and stash as much stuff as you can right now.