The week began with the saddest of news, the sons of Michael Rush getting in touch to let us know that their father had passed away. If you’ve ever clicked around the Rush Database, you’ll have some idea not only of the breadth of his research, but also his generosity. When we met him, Michael was long retired but unstinting in his academic output. By that time he was Emeritus Professor of Politics at the University of Exeter, a specialist in - amongst a great many other things - the social and educational background of Members of the House of Commons. In the course of his research he compiled a dataset covering every MP since 1832. Being a benevolent type, he shared that database with a host of other academics and academic institutions. Amongst them, the House of Commons Library and the History of Parliament Trust.
When we first came across the data, we wanted to put it on a firmer and more supported footing. We approached Michael with the idea of creating a single, online version of the database, with various levels of access from the casual browser to those with enhanced academic access, as well as editors and administrators. Not only did Michael agree to this, he also agreed to publish the data under a Creative Commons Attribution licence.
Michael was a kind and generous man, not only with his data, but with his time and introductions. Without Michael, Anya and our Michael would probably not have joined the Study of Parliament Group or made many of the friends they’ve met along the way. We’re hoping to share more words on Michael in the not too distant future. Farewell, Michael. You’ll be much missed.
Moving on from the sad news to the merely bad, the long hot summer over, Sunday night saw our Members pack school bags and sharpen pencils before boarding the Hogwart’s Express back to Westminster. Which meant our crack team of librarians woke up on Monday morning to an absolute deluge as the floodgates opened and all 3363 written questions tabled in the Commons over recess poured through. Poor librarians. ALL HANDS ON DECK, shouted Librarian Anya, as her crack team wiped the sleep from their eyes and popped open their first can of breakfast Red Bull. Happily, many hands made the work lighter, and all questions were subject indexed in record quick time. Helped of course by our Jianhan’s fairly recent triplestore upgrade [weeknotes passim]. Fine work librarians.
The trials and indeed tribulations did not stop there, Monday also seeing an unexpected, albeit small, cabinet reshuffle. Brilliant timing. All her hands being on deck, there was only Anya left to take up the strain. Popping open the manual and firing up MNIS, she went to work, only occasionally lifting her head to curse said manual. We are told the manual needs work. Everything always does.
And reshuffle repercussions continued to, erm, percuss? As we type these words, it is well past tea time on a Friday night and Librarian Phil has just created yet another ‘Government reshuffle’ Trello card.
Parliament chooses to name and indeed identify its sessions according to the well-established formula of start year to end year. When we woke up on Monday, the current session was most definitely ‘2024-25’. At some point during the week, the government made an announcement that the current session would extend into 2026 and, just like that, we found ourselves in the midst of ‘2024-26’. Now librarians and computational ‘experts’ alike concur that naming a thing does not confer identity, strings and things being quite separate. Unfortunately, clerkly colleagues have been known to conflate the signified with the signifier. Which gets us into a whole heap of trouble.
For reasons lost in the mists, the clerkly naming scheme found its way into our data, meaning written question URIs are described by a composite key of the aforementioned session labels and a ‘unique identifier number’, which turns out to be neither a number nor unique. It is at this point that Anya and Michael roll both their eyes and a cigarette. Did Roy Fielding die for this, they ask?
For more understandable reasons, tabled written questions are sourced from one systen and their corresponding answers from a different system. Answers being matched to their initiating questions by means of their ‘unique’ ‘identifiers’. Dear reader, we believe you may spot the problem here. Anya and Conker being off on their holidays, Librarian Jayne took hold of the reins, yet another Trello board was created and Delivery Manager Lydia marshalled the troops. Even as we type, data engineers Victoria, Sadia and Jianhan are hard at work turning Jayne’s cards into actionable SPARQL queries.
Still, we should not complain. Back in the day, sessions were named and identified according to the regnal years they spanned. In this case the end date of the session and the demise of the monarch. At least these days our naming schema relies on only one unknowable thing.
And now some good news; let’s get back to tales of shipping excellent product.
Fans of our beloved Egg Timer™ will be pleased to learn it now comes complete with a scrutiny start date calculator, the user story being:
As a person working for a Minster
I would like the ability to calculate when parliamentary scrutiny should start
So that I am able to plan the laying of an instrument in a timely enough fashion to allow my Minister to exercise the powers delegated to them by primary legislation by the date on which they wish to do this
Not perhaps the snappiest user story, but you know us. Hardly snappy. In truth, our shiny new calculator did not meet with the acclaim we had dreamed of, offers of funding to take it to a global market not exactly flooding in. That said, our, erm, ‘social media and customer engagement manager’ managed to eke out retweets from both Arabella and John. Which is not be sniffed at.
In other pleasing - if slightly embarrassing - news, our beloved Egg Timer™ was also in receipt of its first ever pull request. At least, its first ever pull request that wasn’t from us. Somewhat embarrassing as Michael had failed to link our day count form to the correct user guides. Pleasing because Matthew was quick to spot the problem and equally quick to issue a fix. Thanks Matthew.
We’d like to thank Arabella, Iona and Ben for the time they put into testing our efforts and the feedback they gave. All of which improved matters enormously.
Jayne, Michael and Robert continue to pore over the notes gathered in recent testing sessions of our Procedure Browsable Space™. Thanks be to Matthews Waddington and Korris. This week saw the main navigation bar further simplified, pagination added to the business item list for a given step and the list of enabling legislation ordered alphabetically. We now feel we are one - maybe two - sessions away from something that might be called useable. And production ready. Whatever that means.
Down on the south coast, Silver and colleagues have made many an improvement to our Single Subject View of the Library™ / Subject Specialist Finder™ / Library Knowledge Base™ prototype. The changes cover everything from data model to data ingest to website serialisation. As some of that data contains contact details for researchers, we’re not able to point you at the results. You’re just gonna have to trust us on this one.
Fans of spreadsheets will no doubt be delighted to learn that yet another in our series of Parliamentary Facts and Figures has been lowered off the maintenance ramp and gained a new MOT stamp from our crack team of librarians and data mechanics. This one covering urgent questions in the House of Commons since 1997. Quite the moving target.
For those of a more pixel persuasion, two more of our applications left the corporate colours paint shop this week. Welcome to your new look UK General Elections Results 1832 - 2019 and Written Statements and Answers. If you prefer your hypertext in suffragette colours, why not take a look.