Understanding Nothing

I’ve recently added the Semantic Mediawiki extension to the Nigov wiki. This is a plugin that allows for the creation of semantic information within a wiki.

The basic principle is extremely simple. Rather than just creating a normal wiki link from one article to another, you annotate that link with information what sort of relationship it represents. This is similar to XFN, and other microformats, except they live in wikitext rather than HTML. Being a wiki, however, they don’t need to be predefined anywhere – the first time you create such a relationship it springs into being as a new “blank” relationship that you can describe or define further.

So, for example, on Nigov we store minutes from Belfast City Council meetings. One of the advantages of having them in a wiki, over the official Word documents, is the linking. The list of councillors who are in attendance can be linked to each of their individual pages. When minutes of previous meetings are approved, they can be linked to. With Semantic Mediawiki we can go even further. In a normal wiki there is no distinction between the link to a councillor who is in attendance, and the link to one who sent apologies. Now, however, we can annotate the links to differentiate between them. On the text of the wiki page for the minutes there is no obvious difference (other than the summary box at the bottom that shows all the semantic information known), but we now have two extra benefits.

Firstly, every page is now addressable as RDF. These can then be fed into a reasoner, or picked up by a semantic search engine. We’re in the early days of the semantic web so this isn’t really as useful yet as it will be in a few years time. But even now there’s still a major benefit: the internal query facility. As well as a built in “Simple Semantic Search”, you can also dynamically include the results of queries into other pages.

So, by annotating each councillor with links like: “[[Is councillor for::Belfast City Council]]”, we can now auto-generate the list of all councillors. Of course, we were always able to do this if we added them to a Category, but if we extend this with “[[electoral area:=Pottinger]]”. and “[[party member of::DUP]]”, we can pull that data into our resulting table as well.

Thus, our list of councillors page, rather than having to be maintained by hand, is as simple as just “ask”ing:

<ask format=”table”>
[[Is councillor for::Belfast City Council]]
[[electoral area:=*]]
[[party member of::*]]
</ask>

Most of the data in the Nigov wiki still needs to be annotated like this before we’ll be able to generate the interesting reports (how often do councillors attend meetings vs send apologies?), and we probably won’t be able to ask the really interesting questions (“have any councillors attended a meeting that allocated funds to an organisation they’re involved in where they didn’t declare a conflict of interest?”) until there’s a proper reasoning engine underneath, rather than the current SQL model.

But for a little extra work when adding links to a page, we get a pretty good payoff now as well as the potential of even more later without having to do any additional work then. Seems like a good deal to me.

Belfast City Council have launched their new website. At first glance it certainly appears to be much better. The one part of it I use the most, the Minutes system is certainly improved, in so far as I can now actually use it again!

The ‘search’ is still fundamentally broken in all the ways I’ve enumerated previously (but at least they contacted me recently to say that they’d put out a tender for a replacement system incorporating a lot of the suggestions I made).

However, they have finally added a simple ‘browse‘ to avoid me even needing to go near the search at all. It’s still rather confusing in that most of February’s minutes, for example, are on the March page, rather than the February page, due to everything being arranged by “council month” rather than “calendar month”, and the browse only goes back to 2002, whereas the search goes to 1989, but it’s still a dramatic improvement, particularly as it works in Firefox.

Here’s hoping that the new minutes system will actually live up to the tender specification…

I’ve received a response from the Council to my enquiries about their broken minutes system. They assure me that nothing has changed with the system, which is not just puzzling, but also means that they have nothing that they can correct. So it appears I’m stuck with a very convoluted way of getting the minutes.

However, they also assured me that they aren’t happy with the current system and lodged a request with the supplier concerning accessibility and browser compatibility. But, as the supplier was unable to give them any firm indications as to when this might happen, the council has decided to investigate procuring a new system.

This will, no doubt, take a considerable period of time to actually happen, but at least there’s some hope for the future.

The nigov site has been rather lacking in updates to the Belfast City Council minutes section for a while now. Although partially this is down to laziness on my part, the main reason is that the Council seem to have changed their on-line minutes system slightly.

Not content with the plethora of problems previously discussed, the latest version of the software spectacularly fails to work with Firefox. You can still run a search (with all the normal problems that entails), but now, when the results are returned, you can no longer paginate through them, or even download the actual minutes.

None of the links on the page are actually true links – they’re all nasty JavaScript links (the next page is “javascript:lisObj._changePage(‘listObj$75′,2)”, for example). And, of course, the JavaScript doesn’t work in Firefox. I wasn’t sure at first whether this is because they’ve updated the software they use, or whether a Firefox update had changed something, but I investigated on a variety of machines with several different Mozilla-based browsers of various vintages, and it seems to be broken on all of them.

I actually forced myself to investigate using Internet Explorer to do this, but although the JavaScript does actually work there, IE’s “load a dummy page for download instructions” mechanism, coupled with the severely broken ‘back’ function of the Council minutes system (which seems to use frames for no apparent reason) means that I can only download one set of minutes at a time before having to return to the home page and re-run the search. Thus downloading a month’s worth of minutes would take an hour rather than 5 minutes.

I asked the Council about this three weeks ago (and whether they have considered the implications of such a broken system on disability access / screen-reading software etc.), but haven’t received any response yet.

Hopefully they fix this soon and I can bring everything up to date again.

In June of last year, a report was provided to the Belfast City Council Policy and Resources (Members) Sub-Committee on the Pilot Study that had been carried out to investigate providing information to the Councillors electronically. Belfast City Council uses over 12.5 million sheets of copying paper per year, and a large number of these are in providing the information packs of minutes and reports to councillors for each Council, Committee and Sub-Committee meeting during the year. It was hoped that moving to an electronic distribution system might help this. (Of course the most likely outcome is that the councillors would just print the information themselves anyway, so the environmental argument probably doesn’t really hold that much weight here).

Unfortunately, the report revealed that the system had mostly been a failure. The website providing the papers was only made available from within the Council’s network, and still required multiple passwords to access. Even though it was ‘internal’, the system was very slow, and often completely unavailable, apparently due to firewall issues. The information wasn’t up-to-date, and there was no diary facility or even a schedule of meetings. As a result, some of the pilot group actually abandoned the system during the trial period

I was curious as to what exactly made the system so bad, so I made a Freedom of Information request for the manual that had been created to explain how to use the system. The system seems to have been built on Microsoft Sharepoint as a fairly straightforward Shared Workspace system, providing the majority (all?) documents in Microsoft Word format. Assuming the documentation fairly reflects the actual system, there doesn’t really appear to be anything particularly difficult about it.

I have to assume, therefore that the main problems were actually to do with the unreliable access, and the difficulties of actually accessing the system even when it was working.

The cover letter that arrived with the User Guide provides another interesting view on this, in that the Council decided that they should redact the Guide before sending it to me, in order to remove the references to the server name on which the system was running. Thus, in the Guide I received, the URL bar in each Internet Explorer screenshot has been blanked out.

The Council have claimed a Section 43 exemption for this, which permits a public authority not to disclose information if its disclosure would, or would be likely to, prejudice the commercial interests of any person, including the public authority holding it. They explained that providing “access to information regarding the server name would increase the risk of the security of our internal electronic systems being breached and would therefore be likely to adversely affect the commercial interests of the Council.”

The Act also requires that, even where an authority believes this, they also have to weigh it up against the public interest in disclosure. As such, the Council goes on to say that although disclosure would allow for a more informed public debate on the issue, promote accountability and transparency, and assist the public to understand and challenge Council decisions, they decided that “the public interest is not best served by putting at risk the security of our internal electronic systems.”

Although I have no particular desire to know the name of the internal server hosting this system, or the URLs of any part of the system, I find the Council’s reasoning here to be rather worrying.

Perhaps it’s just an overly paranoid approach to systems security – and the Councillors’ complaints about how the system could only be accessed from within the Council’s own network, and only then with multiple levels of passwords, certainly bear this out. But I was under the impression that it was generally accepted that security through obscurity is a bad policy.

I’m not going to appeal this one (if anyone else feels strongly about it, I’ll certainly give them a copy of the letter and let them take it up with the Council), but I think I am going to make a follow-up request for more information on the information gathered from the Councillors about the trial system. As with other things I’ve discovered in the Council minutes, I support the ideas and ideals that the Council appear to be espousing with such a system, but I think the actual execution leaves a lot to be desired.

Action Renewables is a joint venture between the DETI and Viridian to stimulate awareness and discussion of renewable energy technologies.

In 2004 they had an excellent plan to help raise awareness – power the lights of the Belfast City Hall Christmas tree using a wind-powered generator.

Everyone seems to have thought this was a good idea: the Council gets to show its committment to promoting renewable engergy, Action Renewables gets some excellent PR, and the public gets to see that renewable energy in action.

Unfortunately it seems that, much to the surprise of everyone else involved, Planning Services decided that, as the City Hall is a Listed Building, planning permission would be required for the temporary generator. And so, although the Council agreed in principle to the idea, it had to be shelved for a year to allow everyone to properly work through the details of the proposal.

However, I couldn’t find any record of this actually being brought back before the Council again this year, and so made a request to find out what had happened. The Council have confirmed that this wasn’t actually brought back before them this year, and have suggested contacting Action Renewables’ PR firm to find out what happened.

My suspicion is that this was simply forgotten about, so it’ll be interesting to see how they respond…

I’ve just been informed by Food Safety in Belfast City Council that they have plans to publish their inspection reports on the web, but are currently awaiting legal opinion to make sure it’s OK for them to volunteer this information openly (rather than as a result to a specific FoI request).

Assuming they get the go ahead they hope to have a site live in the first half of 2006 which can then be used as a pilot site so that all councils across the UK can present the information in a consistent manner.

Whilst I await the detailed response to my request, I should probably make a similar one from several of the other local councils, as lots of the restaurants I eat at semi-regularly are outside Belfast…

When I’ve visited Boston in the past I’ve always liked that the local press publish accounts of food safety inspections, and have often wondered why we don’t get that here.

I suspect that that’s about to change. The Information Commissioner has ruled that local councils have to make public their reports on the health and safety of restaurants they inspect.

No doubt it will take Belfast City Council quite some time to get around to implementing a system for doing this, so to kick things off I’ve made an FoI request for details of all inspections in the last two years.

It seems that Belfast City Council agree that their website is terrible.

They’re worried that traffic to their website has declined (although I suspect they don’t really measure this very well), and also that the recent report by the Society of Information Technology Managers on the website of all UK local authorities gives them bad marks compared to the other major cities in the UK (and even some of the other councils in NI).

So, they’re going to do something about it.

They’re not going to fix any of the technology though. They’re just going to hire a team of 3 Web Editors for a 12 month period to write better content for the site.

Somehow I doubt that this will help much …

Yesterday evening, whilst processing the minutes of the Belfast City Council Policy and Resources Committee from 18 March 2005, I noticed that they referred to a report into the Council’s strategic approach to communications – including a review of the Council’s corporate identity, the development of the website as a key communications tool and a review of the Council’s approach to graphic design, advertising, publications and leaflets. This sounded interesting, but, unlike most reports presented to the committees, this one wasn’t included with the minutes. So, at about 5:30 yesterday evening, I sent off a Freedom of Information request for it.

At about 10:30 this morning, I received a copy of it in email! This is probably the fastest result of a FoI request I’ve had yet!

If anyone else finds it interesting, I’ve added the report to the nigov wiki.

The section on the City Council website is a little bland:

Belfast City Council’s website needs to be properly resourced and positioned so that it plays its part in disseminating the council’s key messages. It also needs developed in terms of providing information and, even more importantly, services to ratepayers. The website is vital to the council’s reputation not only at home but overseas. A proposal is going before P&R to properly resource this operation and it needs to work closely or as part of Corporate Communications to ensure opportunities are maximised.

I’ve sent a follow-up request for the proposal.