2020 - Week 45

As we re-entered lockdown, week 45 was more meeting heavy than anyone had any right to expect. And yet, despite having faces glued to pixels and poor Robert’s ears being glued to misfiring headphones, it also proved remarkably productive.

Modelling their Lordships

On Wednesday we took a rare jaunt to our beloved palace. Anya, Donna, Ben, Silver, Robert and Michael kept a respectful distance in the House of Lords Journal Office and set about sketching what we think we know about Lords’ memberships and how they come in to being. Unfortunately Ben had forgotten to arrange a whiteboard, or anything vaguely resembling a whiteboard. So we grabbed some paper and, taking great care to not get Sharpie stains on the carpet, made sketches of hereditary peers, excepted hereditary peers, life peers, Law Lords, hierarchical bishops and named bishops. We think we got to a fairly comprehensive model of seats and incumbencies and the virtues by which noble bottoms fill them. We still have some questions. Not least around Law Lords and the mechanics of inheritance. We’ve been taking David’s peerage database as a reasonable start on the non-bishop part of this, so a missive has been dispatched and we hope to have a better understanding of how remainders and special remainders power the waterfall of privilege. Or at least how they used to.

In the short term Robert and Michael plan to fold their learnings back into the database they purloined from David. In the longer term this should feed into any solution for modelling Lords’ incumbencies, and hopefully help us get a better understanding of what went wrong with orchestration into the data platform. The model used at source has no notion of Lords’ seats or what causes a person to occupy one. And at some point some software was built to do a thing called ‘Lords’ Inflation’. And at some point that stopped working. Nobody is quite sure why, but there’s a suspicion that it might have had something to do with ex-colleagues leaving the building. Silver and his friends at Data Language continue to investigate.

Procedure mapping

On Thursday, Arabella got in touch to ask whether it would be possible to add new steps to the treaty procedure data, to describe a Commons select committee opening an inquiry into a treaty. To date, we’ve represented what happens when a committee reports on an inquiry. She wondered whether we might perhaps add new steps for the Environment, Food and Rural Affairs Committee, which has just opened an inquiry into the fisheries agreement with Norway. And whether we might also add steps to cover Government responses to committee reports. So on Friday, Librarian Jayne and Michael sat down with marker pens and erasers and, with the usual diligence, once again attacked the treaty map. Which really does start to look like a spider’s had a stroke. On the plus side, the Norway fisheries treaty page now shows steps and links for the opening of the EFRA inquiry and the UK / Japan treaty page does the same but for the International Trade Committee inquiry. We still plan to be reactive rather than proactive when it comes to adding Commons’ select committees, only adding them to the maps as and when they choose to scrutinise treaties. Not least to save on ink. The question of how we’ll find out when a committee has picked up a treaty remains unresolved. If anyone from select committee land is reading and has an answer to this we’d be delighted to hear from you.

This week also saw a couple of those tickets where the correct response turns out to be “do nothing”. The first was occasioned by the Government making a written statement setting out proposals for the scrutiny of the free trade agreement with Japan. We had wondered if this was something that should appear as part of the treaty procedure but, after checking with Alex and Dominique, the answer turned out to be no. Neither of our legal colleagues considered the statement to be procedural in nature. This came as something of a relief. How one might add something that predates laying to a procedure where the procedure starts with laying being a problem that neither librarians nor computational experts really want to grapple with.

Written statements returned to the fore when the Minister for Business and Industry made one to let it be known that the latest ‘Rona Restrictions would come into force before being laid before Parliament. We checked in with JO Jane, who stated that the statement was not procedural and should not be included. We’re already covered for querying for such things as both our made negative and made affirmative maps include steps for statutory instruments coming into operation before being laid. These being required by House of Commons public Standing Order 160 and House of Lords public Standing Order 71. So most definitely procedural.

Your weekly egg timer update

If you’ve been following along from home you’ll know that, way back in the mists of pre-lockdown times, JO Jane presented us with a Word document outlining the calculus she uses to generate anticipated scrutiny end dates for assorted types of instrument. Being nothing if not optimistic, computational expert Michael thought he’d take the document and turn it into code. Which felt at the time like a few hours work. Or a ‘one bottle of wine problem’, as he likes to say.

After several months of confusion - not least about what Jane might have meant by ‘bums on seats’ - and many hours of staring at legislation with no formal education in what the special words mean, and a small crate of red, this week the egg timer of parliamentary scrutiny finally ‘went live’. Jane did some testing and declared herself happy. And the Twitter based community of CRAG-heads also appeared content. Being gentlemen of a certain age, Robert and Michael remain convinced that working in the open does indeed make things better. And Arabella seems to agree.

That said, before the bunting went up and corks were popped, there was still work to do. Monday morning saw Robert and Michael pouring over the Legislative and Regulatory Reform Act 2006 section 19, the Public Bodies Act 2011 section 11 (12) and the Localism Act 2011 section 19 (14), in the hope of finding what flavour of days should be counted. Luckily the acts concur with each other, bringing the count to a halt when either House is adjourned for more than four days. Robert and Michael already had a special case made affirmative calculation dealing with this logic, so it only remained to add comments to cover new cases, and the job was done.

On the subject of comments, you might know that Robert and Michael are rather proud of theirs. They’ve probably spent more time writing comments than they have writing code. All of their comments are written in Markdown with links to cite informing legislation. Robert has even written a parser that takes their code and makes it into more Markdown and, of course, HTML. Which means the logic behind the calculations should be understandable by anyone. Or at least anyone with an interest. Which is probably a fairly small subset of anyone. The HTML generated by the Markdown parser had one small problem. Long lines of code were not being properly wrapped, so any error messages and suchlike were disappearing off all but the largest of screens. Luckily our Samu saw the link pop by on Twitter, clocked the problem and put in a pull request. Which fixed our line lengths nicely. Thanks Samu.

At some point in the week, Michael noticed that the treaty calculation had gone somewhat awry, invoking an odd mix of parliamentary sitting days and actual sitting days. Unless and until one or other House sits through the night and into the next day, the date returned would be quite correct. But we haven’t come this far to give up on pedantry now. So the error was fixed. And with that, Robert somehow made the date input work on a tiny screen and added an egg timer favicon. The icing having gained its cherry, we pushed live.

The resulting tweets included a call for any treaty heads to check our maths. Should they have time. Arabella kindly volunteered. So on Friday, we returned to Teams and talked her through the code and the comments. She pointed out that - whilst the start date we ask for is the laying date, or, for period B, the date of any ministerial statement - the actual start date of scrutiny is the first joint actual sitting day following laying. There’s not an awful lot we can do about this. Asking the user to input the scrutiny start date would require them to have knowledge of sitting days and run a calculation to get the start date before they could run our calculation to get the end date. But Robert and Michael have now updated the code to output the scrutiny period start and end dates, as well as some indication of the nature of the calculation start date. Which has made us wonder if the various clock start steps we maintain across procedural maps have been inputted correctly. Legally speaking. A question for next week.

Glitching all the things

Young Robert continued his glitch tinkering, splitting his time between Lords business tidies and making fragments of legislation more easily citable. If you visit our favourite section of the Statutory Instrument Act 1946, you can now toggle between whether you want to see fragment identifiers - “ID badges” - exposed. Or not. And if you choose to have them show, you’ll notice they’re now accompanied by a plus button. Should you click on said button, the full path to that fragment on legislation.gov.uk will - by some process comparable to magic - be copied to your clipboard. Wow, Robert.

Not only that, our Robert has also written a bookmarklet. Just visit his legislation application, pop down to the bottom of the page and drag it to your bookmarks bar. Next time you’re looking at something on legislation.gov.uk and think to yourself, “My word, I’d like to link to that bit but I can barely be bothered fighting my way through the HTML source”, just click the bookmarklet, click that plus button and you’re good to make links. And not only that, once you’ve copied the link you want, merely click the bookmarklet again and you’ll be back in the comforting arms of TNA’s finest website. Michael, for one, shall be making full use of this.

Tidying up behind ourselves

Since we first stumbled into the world of statutory instrument procedure not only have our lives improved immeasurably, our appreciation of legislation.gov.uk has also improved. Also immeasurably. Over the past couple of years, we’ve made a lot of links. A lot of links. Some of those links were to http. Since we went live with the SI service and Treaty tracker, TNA have made the shift to https. Obviously, redirects are in place and nothing breaks. But it’s always better to store URLs as they are and not as they were. So this week, our Jianhan updated all our links which had been to http://legislation.gov.uk to https instead. Because we are librarians at heart. And therefore tidy. Very tidy.

All about the collaboration

Way back on Monday, computational experts Robert and Michael joined a meeting with Anya and Liz at which Librarian Ned talked through his work on linking organisations appearing in Parliament’s thesaurus to their equivalent concepts in Wikidata. And vice versa. Some time back we dipped toes into the waters of cross-linking our terms with theirs. But not at the time of creation. Ned has been looking at how we might do the interlinking as we add new organisations - at the point of creation - what’s tricky about it and also what benefits that might bring. All very fine work. They also chatted about earlier - Andrew powered - efforts to ease interlinking and how we might pick that work up. Because we are nothing if not native to a web of data.

Way back at the start of these notes - mere minutes to our reader, months to our authors - we mentioned our trip to the House of Lords Journal Office. Not only did we not draw on the carpet, we also took the time to have another peep around the Hanging Files. More of which we covered back in week 44. Since then we’ve had multiple meetings with Lords clerks, to chat about how they use the files, what works and what doesn’t. We’d like to thank Luke, James and Duncan for their help. If we’re yet to chat to you, we basically have three questions:

Lifelong learning

On Thursday evening a decent selection of our team of crack librarians went along to a lecture by Vanda on the principles underlying Knowledge Organization Systems, where systems of classification enumerative, analytico-synthetic and faceted were the order of the day. Your correspondent will not be taking questions at this time.