Government domains update

I blogged a little while ago about the rather bizarre situation whereby neither the government nor their domain registrars were willing to tell us which government domains had been registered. In that post, I mentioned another FoI request that I’d sent which attempted to obtain the information via a different route.

Possibly as a result of this, it would appear that sanity has prevailed and the government has now published a full list of domains. You can download the list in csv format from the Cabinet Office website.

I’m still a little intrigued, though. The original refusal by JANET (the academic and government domain registry) to release the list was justified on the grounds that release of the material in full is considered to “present a risk to the commercial interests” of JANET’s customers. Now that we have the list, is there anything in there which might support such a view? Are there any domains which could be considered commercially sensitive? If so, which ones, and why? And why was JANET so determined not to release the list in the first place?

Update: I’ve taken the list of domains, put them in a database and added some meta-data about them (sourced from whois and DNS). You can see the results at

Why aren’t we allowed to know which domains the government has registered?

If you’ve already read this, you might want to skip to the end where I’ve added some extra information.

You’d think, wouldn’t you, that there ought to be no problem finding out which official domain names (ie, those ending in have been registered by government departments and organisations at both national and local level. It is, after all, one of things that, even if it isn’t directly published anywhere, should be amenable to a simple Freedom of Information request.

Unfortunately, this isn’t the case. And the various excuses given for not revealing the list are, to say the least, increasingly bizarre.

The story starts with an FoI request by John Cross to the Central Office of Information (COI) for a list of domain names ending in The COI responded, not unreasonably, that they don’t hold the information themselves. However, they noted that it could be obtained from JANET, the organistion respondible for registering domains. Other people have made similar requests to COI, and got the same answer.

John didn’t immediately follow this up himself, at least not via, but in April 2010 a similar request was sent to The JNT Association (aka JANET) by Steve Elibank. Entirely unreasonably in this case, the request was refused. JANET gave two reasons for their refusal:

1. The material is already available via the whois service (refusal under section 21 of the FOIA),
2. Release of the material in full is considered to “present a risk to the commercial interests” of JANET’s customers (refusal under section 43 of the FOIA).

As was pointed out in annotations to that request, these two reasons are mutually incompatible – if it’s already available, then releasing it can’t be a commercial problem. More relevantly, it simply isn’t true that the material is already available – whois will give you information about domains that you know about, but isn’t helpful for discovering the existance of domains that you don’t know about.

Not to be deterred, David Batley made a similar request in April this year. This, too, was refused on exactly the same grounds.

At this point, John Cross re-entered the fray with another request in which he detailed the reasons why neither section 21 nor section 43 were valid reasons to refuse the request. It’s also worth noting here that other requests for a full list of domains (ie, those owned by further educational establishments) and domains had been successful.

After much to-ing and fro-ing, JANET eventually responded with the answer that, as a list of central government websites was online at the Cabinet Office website, they were refusing the request under section 21 (maerial available elsewhere). John pointed out that this wasn’t an answer to his question, and in the light of JANET’s continuing refusal to answer it, sent a formal complaint to the ICO. We still await the outcome of that (which may take several months, the ICO is not noted for the rapidity of its investigations).

That’s where I got involved. I was intrigued enough to wonder why JANET didn’t want to release the full list, so I submitted an FoI request to try to get some background information. In particular, I wanted to see the wording of the advice given that releasing the list would be a risk to commercial interests, and also to see if they had any documentation which explained why a list of domains can be released but not a list of domains. JANET’s response to this one, so far, has been to simply ignore it. They are now beyond the legal maximim period of time in which to answer FoI requests, but I’ve still got nothing.

More intriguingly, though, while trawling through the COI website, I came across the minutes of a meeting of the Naming and Approvals Committee in 2009. These contained a couple of action points:

JANET to notify COI of all domains for renewal at least 3 months in advance.


Review current list of domain names to check for redirects

Assuming these action points to have been, well, actioned, this rather gives the lie to the original claim by the COI that they don’t have the data as the committee cannot have carried out the second without it and, since this started in 2009, the first will have given them a full list by now (although, to be fair, this is a rather obscure reference and it’s entirely possible that the staff member who gave the original response was simply unaware that the committee might have the data). What I’ve now done, therefore, is go back to the COI with a new FoI request to see the contents of these notifications and the list. They’ve got until the 3rd October to answer that, so we’ll have to wait and see how that pans out.

In the meantime, John Cross has submitted yet another FoI request to JANET, this time asking just for a list of domains which aren’t included in the list available on the Cabinet Office website. That was sent on the 17th August. It’s now the 7th September, and the response is due no later than the 15th of this month. Despite that, John hasn’t even had the courtesy of a acknowledgement from JANET.

I would really like to know what’s going on here. Why is a list of government domain names considered commercially sensitive, when a list of educational and police domain names isn’t? Why has JANET suddenly become so unresponsive? Does anyone outside JANET have a clue? Does anyone inside JANET have a clue?

Update: A couple of things have transpired since I wrote the above. Firstly, by making enquiries elsewhere I discovered that JANET had, in fact, replied to these requests but the replies hadn’t got through – because they’d sent them with the destination address in the Bcc: header and nothing in the To: header, meaning that it inevitably got caught by spam filters. After having a certain amount of clue applied, the responses were re-sent. You can follow the link back to the request if you want, but here’s the response in full:

Dear Mark

The legal position is that JANET(UK) is not a body that is required to respond to requests under the Freedom of Information Act. We have helped as far as we can and will provide no further information on this request.

Tim Kidd
Operations Director

The same response has been sent to John Cross here and here.

Unfortunately, this statement is contradicted by JANET’s own documentation. Their publication scheme states:

The Freedom of Information Act received Royal Assent on 30 November 2000. The Act requires all public bodies to adopt and maintain a publication scheme. JANET(UK) is a public body for the purposes of this legislation, since it is a company wholly owned by bodies that are public authorities viz. the UK education funding councils.

Oddly enough, that document is no longer linked from the website. Just in case they decide to remove it, I’ve kept a copy here.

So, what do we make of this? I’d really like to know more. If anyone inside JANET wants to be a mole, then please get in touch.

Points mean….?

I don’t normally go in  for the Disgusted of Tunbridge Wells approach, but the answer to this Freedom of Information request very nearly had me heading into Daily Mail territory with my reaction.

The question asked was fairly simple: The inquirer wanted to know how many people are entitled to drive (ie, are not banned from  driving) but have 12 or more points on their licence – that is, have more than the number of points which would normally result in an automatic ban. The answer is shown in this table.

Now, I’m not too bothered about the people with 12 points, or numbers a little over that. The law does allow for discretion when it comes to imposing a ban in cases where it would cause “exceptional hardship”. And it’s quite easy to reach 12 points without ever doing anything seriously wrong –  a couple of fixed-penalty speeding tickets issued by cameras followed by a single instance of driving without insurance as a result of a genuine oversight in missing the renewal date by a day, for example, would do it. So there are probably quite a lot of people who reach 12 points and can reasonably count themselves unlucky to be banned, as well as a fair number who are deservedly allowed to carry on driving.

But when you get to the higher values – at the bottom end of that table – then the mind does start to boggle a bit. There are people driving around with more than 20 points on their licence. To get that, they’d have to have committed multiple minor offences, or at least one serious one, after reaching the point at which they would normally be banned.

And, right at the end, we discover that there is (or, at least, on the 7th July 2011 there was) someone in the UK in possession of a perfectly valid driving licence with 33 points on it. That’s nearly three times the number which normally results in a ban. It’s 11 speeding tickets, or three convictions for dangerous driving. So the question has to be asked: Just how bad a driver do you have to be to rack up that many points? How much have they paid in fines each time? How big a hardship would it be for them to be banned? And how much does their insurance cost? Whoever it is, they must be getting pretty close to the point at which it would be cheaper to take a taxi every time they want to go somewhere.

Unseen side of justice – further witterings

My previous post on the FOI request about Courtel and HMCTS has already started doing the rounds of Twitter, Facebook and the blogosphere, so I thought I ought to clarify something that maybe wasn’t plain before. I don’t object to Courtel making money from the data – their Courtserve 2000 software (and even the email alert system) is clearly a value added service and they’re fully entitled to charge for it. What I do object to is the fact that Courtel have a monopoly on the data and that monopoly not only makes them the only organisation able to exploit it commercially but it also prevents non-profit organisations, community groups and individuals from using the data effectively. I can see no reason why court lists should not be published by the courts themselves on their own website under an Open Government licence – it would cost them no more than they currently spend on giving it away to Courtel, and it would allow a much wider use than is currently possible. If Courtserve 2000 is as good as Courtel say it is, then allowing competitors isn’t going to hurt them too much. On the other hand, if it’s not as good as it’s claimed to be, then some competition will be in everyone’s interest.

On a separate note, @johnlsheridan on Twitter has alerted me to section 14 of The Re-use of Public Sector Information Regulations 2005, which reads

Prohibition of exclusive arrangements

14.—(1) Subject to paragraph (2), a public sector body shall not enter into an exclusive arrangement with any person including an applicant.

(2) A public sector body may, where necessary for the provision of a service in the public interest, enter into an exclusive arrangement.

(3) The validity of the reason for granting the exclusive arrangement under paragraph (2) shall be reviewed at least once every three years.

(4) Any exclusive arrangement permitted under paragraph (2) and entered into on or after 31st December 2003 shall be published by the public sector body.

(5) Any exclusive arrangement which exists on the date of entry into force of these Regulations and to which paragraph (2) does not apply shall be terminated at the earlier of—

(a)the date on which it comes to an end in accordance with its terms; or

(b)31st December 2008.

(6) In this regulation, “exclusive arrangement” means a contract or other arrangement granting an exclusive right to re-use a document.

The contract between HMCTS and Courtel appears, on the face of it, to be illegal under this clause unless there are reasons why such an exclusive arrangement is necessary for the provision of a service in the public interest. I would be very interested in seeing their justification for this arrangement.

The unseen side of justice

In R v Sussex Justices, Ex parte McCarthy, the then Lord Chief Justice Lord Hewart delivered a judgment containing a phrase which has passed into common parlance. Commenting on the need for the legal system to be both transparent and open, he said

…it is not merely of some importance but is of fundamental importance that justice should not only be done, but should manifestly and undoubtedly be seen to be done.

You would have thought, then, that in this day of the Internet, open data, freedom of information and open government, the principle that justice should be seen to be done would extend to the publication of relevant material on the web by the court service. Unfortunately, this is not necessarily the case. While some courts (notably the new Supreme Court of the United Kingdom) do routinely publish hearing lists and judgments on their own website, this is the exception rather than the rule.

There’s a whole lot which could be said about this, but for the sake of brevity I’m only going to mention one particular aspect of it. It used to be the case (or so I’m told, I’ve never actually had reason to do it myself) that if you wanted to know which cases were scheduled for any particular county or crown court on a particular day you could phone them up and they’d tell you.  In the pre-Internet era, that was probably the best and most practical option. It was free (well, no more than the cost of a phone call), and it was accessible to anyone: legal processional, journalist or layman alike. So, in the 21st century, the obvious successor to that system would be for the courts to publish their hearing lists on the web so that anyone can find the information for free (well, no more than the cost of your data connection) – and it would still be accessible to everyone, professional and layman alike. There would even be scope for third party organisations to take the data and produce mashups of it (maybe correlate with the crime figures published on, for example, or create an email alert service for particular courts and cases).

So, what has happened? The answer is that it’s nothing like that. Rather than publish the material themselves, the court service has chosen instead to contract it out to a third party. That in itself wouldn’t be a problem if the new service was run on the same lines as the old – open, accessible to everyone, and free. But, as you’ve probably guessed, it isn’t.

The current, and exclusive, providers of crown and county court hearing lists are Courtel Communications Ltd, via their Courtserve website. While a lot of basic information is accessible on their site without charge, a lot more of it requires a subscription. And even much of that which is free still requires registration, so it isn’t possible to simply browse the site and find the information. More importantly, the Courtserve site appears to be aimed exclusively at legal professionals and what they call “authenticated media”. A company name, telephone number and “business email address” are required fields on the registration form for the free services, something which will severely discourage non-professional use. The licence terms on their website explicitly prohibit any copying or redistribution of the data, which makes it impossible for other organisations to legitimately add value or functionality to it. And, finally, I am also reliably informed that the telephone enquiry service has been discontinued – callers are now referred to Courtserve instead.

I was intrigued enough about this to send in a Freedom of Information request to the court service about their relationship with Courtel and Courtserve.  You can read their full response by following that link, but the executive summary is that Courtel, despite charging for access to it by end users, are not paying anything to the court service for the data. Nevertheless, the court service has given them the exclusive right to distribute it (so no other organisation may do so), and this even extends to an explicit commitment by the court service not to publish the data on their own website.

I’m not opposed in principle to privatisation of public services, so long as it delivers benefits to the taxpayer and the consumer. But this arrangement clearly doesn’t. Courtel are getting all the benefits of a data monopoly, while the taxpayer (via the court service) gets nothing in return as the data is being given to them for free. And the consumer loses out as the restrictive nature of Courtserve means that public access to the information is unduly limited and costly. (Not to mention the fact that Courtel’s contract with the court service requires them, if requested, to pass details of their subscribers to the court service. I wonder how many subscribers are actually aware of that?)

I think this is bordering on scandalous. While other departments of government have made great strides towards opening up their data for public use, the Ministry of Justice and the court service have quietly headed off in the other direction and made access to their data less open instead. It’s time that was changed.

Update: I’ve added a follow-up post to this one in the light of some more information and feedback from Twitter.