Three years ago, ID cards were still the obsession of a small number of MPs. Now we are well into the spin cycle.

This weekend we witnessed the familiar pattern of what passes for political debate these days.

Rumours are leaked to the papers about cabinet splits. There's feverish speculation about who's for and who's against. And at the end of it all, a marginal idea has been spun into the mainstream.

That's not an issue for most IT professionals beyond their personal viewpoints. What is important is the gap where discussions about technology ought to be.

Anyone who sees technical issues as mere detail in the ID cards debate is kidding themselves.

Let's start with biometrics. There's been a bit of discussion here, and concerns have been raised over nearly all the increasing number of potential options, from fingerprints to iris scans.

The truth is that few have been tried on a grand scale. Whenever you hear claims for biometrics, there is always reassurance that the chances of it going wrong are one in 10 million (or some other suspiciously round number).

It's an impressive success rate for a limited role, but not when you consider the vast numbers of potential uses every day.

You may recall similarly inflated technology claims - CDs will last a lifetime, for example. And one should also remember the statistical truism about terrorism: the state must be lucky every time, the terrorists just once.

But the real problem, one that is not remotely close to satisfactory resolution, lies with the database. Biometrics is only the front end of any system. The biggest dangers may lie in the database on which they draw.

We must have an open and informed debate about this. Screw up the IT on this project and you potentially increase the very threat you are trying to avoid.

The debate, with all the technical details, is too important for it to become yet another victim of spin.