« i have no words | Main | occupying the SuGaytenland »

November 08, 2011



I'm incapable of feeling any sympathy for Theresa May. She helped create this particular toxic sewer, so it seems reasonable that she should be the one suffering, though I agree with the larger point.

And it's her department that plans to make skilled immigrants leave after five years - an idea which is not only insane, has no real popular support and which is going to cause serious problems for British companies*.

I am fascinated by the sheer political naivety on display, though. Did she think the civil servants would forget about the emails absolving them of responsibility.

* I realise it will also cause huge problems for British universities, but I think the whole student visa issue, the student fees issue, the research funding changes all demonstrate the Tories want to destroy the university system, so that aspect might be a feature.

Charlie W

Perversely, semi-facetiously, and in a reverse of previous thinking, I wonder if the best liberal strategy here is to call for the passport checking problem to be handed over to machines; that is, to the automatic gates with cameras that they have at some UK airports now. The argument goes as follows:

First, the machines will have a false positive and a false negative rate. They must do, since if they deliberate for too long, too many people with the belief that their documentation is legitimate (a category which includes almost every traveller, most likely) will object. Similarly if they call for help too often (this also causes a radical slowdown). Sometimes (often?) the machines will be opening those gates in situations when a better placed credentials assessor, with better information, would decline to open them.

Second, the machines will have no way to assess whether or not they are doing their jobs properly. They will just continue to operate uncomplainingly within the parameters they've been given (I'm ignoring trivial fault reporting).

Third - and crucially - the authorities will believe that the machines are infallible. I suspect they may even be willing to assert that the machines know who's legitimate and who isn't.

Hence, by contrast with the present arrangement involving human passport controllers, we could arrange a situation where the authorities (and the inhabitants) of Essex are unwittingly content to have slightly leaky borders.


And she had the gall to threaten to have civil servants who made decisions she didn't like subjected to criminal charges!, like some Soviet potentate having the area railways manager arrested because a train was late.

I don't think I've ever heard of a UK government minister actually threatening officials with police action over an administrative decision before.

Another story here is the predicted failure of biometrics, of course, as the original problem was that it took longer to process the superduper new thing than it did for an IO to look at it.


Having just gone through Heathrow's passport control a few weeks ago, I had the opportunity to observe the super-duper high-tech biometric bullshit in action -- from the relative safety of the human-only queue. During the time I spent in the queue, I saw that system fail to recognise the person's face far more often than it succeeded, and every time it failed, the individual had to go through to an immigration officer.

On the occasions where it succeeded, the subject would often laugh happily and wave his/her passport as though it had been some sort of video game.

The entire thing was ludicrous.


To recap, it seems that the Home Secretary told the relevant organs not to treat visitors as potential terrorist or criminal suspects. Not because this was wrong for moral or practical reasons, you understand, but because it was causing delays.

By "treating visitors as potential terrorist or criminal suspects" we are talking here about "looking at their passports", correct? It's immoral to look at someone's passport now?


It's immoral to look at someone's passport now?

No. It's just a teensy bit pointless if this chap is to be believed:

"...only about 2,500 people a year are intercepted at the border, and only about half of them are refused entry to the country because they are on the Home Office's warnings index of illegal migrants and potential terrorists.

That is 2,500 out of 125 million – it is like looking for a needle in a haystack."

As for 'handing it over to the machines', I fear we in that over familiar territory of a Large Scale UK Government Computing Contract:

"The Home Office has long pinned hopes of getting rid of the embarrassing passport and security queues at airports on its flagship e-borders programme, and it is now nearly 10 years since work on this much-delayed project first started.

The idea is that all 125 million passengers a year who enter and leave the UK will eventually be logged electronically on entry and exit, thus delivering the holy grail of a secure border.

The system relies on the airlines giving advance notice of passenger details to the UK Border Agency before a flight leaves to travel to Britain so they can be checked against watchlists of potential terrorists or illegal migrants before they get on the plane.

In effect, passport checks are being shifted upstream so the screening takes place at check-in rather than passport control in the airport arrivals hall. Around 68,000 people were stopped coming to Britain this way last year compared with the 2,500 at the border.

... The national border targeting centre, in Salford, has started work storing this data to build up travel histories of every logged passenger.

But progress is slow, and only 55% of the passengers who come into Britain are actually screened in this way. The e-borders project itself is mired in a complicated legal dispute, with the previous contractor, Raytheon, and the Home Office counter-suing each other in a £500m arbitration claim."

chris williams

I could have told them that any attempt to mechanise the Traffic Index, even with a high degree of preregistration through electronically readable passports, would be a very difficult task. Thing is, that's what they found out in 1959.


To be fair, Chris, technology has moved on a little from 1959.


1959? That's before most of the people probably involved in the modern fiasco were born. Institutional memory and professional histories of public services are sooo last century.


I think there is a bit of a problem with the assumption that, with no passport controls at all, there would still only be 2,500 people a year trying to get into the UK who shouldn't be. It's not like Geert Wilders and Fred Phelps keep trying again and again.


There's an argument to be made for open borders, but I don't recall anyone involved in the story in question, including Alan Travis, advocating that. So the idea of 'no passport controls at all' seems a bit of a straw man.


NZ has a system in place similar to that described in CMcM's quote. It was brought in during the mid-90's, IIRC, and I had the unfortunate experience of working on it for a short while.

The whole thing was very shaky. Unsurprisingly, considering Anderson Consulting (as they were) had developed it with their usual skill, understanding and careful attention to detail.

I think it could best be summed up by noting that I was the sole programmer responsible for maintenance of the entire server-side system, I was 22, and it was my first professional job.


Ajay: the point of the suspect (or warning) index isn't to check everyone's passports against it. It's to check the people who you *suspect* against it. Suspects, who are already known to the police, are not going to travel on their own passports if they can help it.

It was a stupid idea in the first place.


Duaneg, isn't that how Accenture typically works. The most junior hires (often English, or history graduates, bizarrely) do the programming, the next layer manage them, the next layer manage the managers... They seem to rely upon contractors for the technical stuff. I always used to find the conversations with their developers alarming, as they seemed to have no idea what they were doing.

One day somebody must write a book about software gets written in Accenture, if only to explain what the hell it is they do.


>>>"And the opposition, meanwhile, is making an utter disgrace of itself hopping from foot to foot"

I've not been following this one particularly carefully and my main reaction so far to random snippets heard off the radio has been "Oh, fuck off, Yvette Cooper".

With my track record of divergence from the political choices of the British people - I give you the next Labour Prime Minister.

Seriously, genuine question: how does Yvette play with the focus groups? (Or has that approach just kicked the bucket?)


Yes, I understand this was (is?) their typical MO. I was a short-term contractor, as were the other coders and assorted techie types, from what I recall.

The permie AC people were manager types and were generally awful. It was obvious they viewed the project as a tedious stepping stone to better things, and they were openly contemptuous of the remaining customs & immigration IT staff.

It was frightening and infuriating. These assholes were supposedly responsible for the systems that were protecting our borders and they were clearly incapable of and uninterested in doing the job. Furthermore they had totally alienated and mostly driven out the people that were.

They certainly made a lot of money, though.

The thing that really gets me is that as far as I can see all big consulting jobs are like that, including all big government IT projects. NHS, military, whatever. It makes me quite angry.


Could be, I'm not sure. Accenture are in a special, and fairly horrific, category of their own though. Other consultancy firms are often incompetent, don't have good people (because if you're good - why would you stay?), are as you say indifferent to their projects and tend to be very bureacratic. The techniques they use tend to be from the stone age.

Accenture on the other hand seem to have gone out of their way to be really terrible. Its like they carried out research into the worst possible approach to software development and then built it into their corporate DNA.

I paid a lot of attention to the NHS project, partly out of professional interest and partly because I knew people working on it. It made me very depressed. Fucking up because a problem is hard is one thing. Fucking up because you're using methods and techniques that were discredited 20 years ago is a completely different thing. There is a whole literature on medical record systems, with case studies and everything. To just ignore that... criminal.

The comments to this entry are closed.

friends blogs


Blog powered by Typepad

my former home