2022 - Week 38

Again, again, again, we are forced to kick off with apologies. When we started writing these notes it was with the very best of intentions. Quick weekly updates of what we’d done, what we’d failed to do and what we’d learned. Several years later and our poor notes appear to taken on the cadence of sessional returns. And the length. We can only blame deployment pipelines, our own recalcitrance and events. And what a series of events. Since we last spoke, everyone’s favourite team of crack librarians have been dealing with resignations, appointments, a new Prime Minister, further appointments and a change of monarch.

At this point we should single out Librarian Anna for her well deserved star award, having performed a sterling job of manning the MNIS wheel in the midst of an unexpected slew of ministerial departures. And subsequent appointments. Yes, it does feel like ages ago. Top work Anna.

And with that, back to business. Read on, dear reader, read on.

Breaking egg timer news

One consequence of recent events has been a whole series of unexpected changes to sitting days. Once again, our crack team of librarians have stepped up to the mark, managing both the Google calendars that sit behind our beloved egg timer and reassigning scrutiny end dates to instruments, both statutory and treaty-like in nature. Applause to Librarians Ayesha, Jayshree and Martin for banging egg timing matters to rights.

The application has been running on Michael’s Heroku account for the past three - has it really been that long? - years, at a cost to his purse of close on twenty Great British pounds a month. Given the current financial climate, this was not a cost he could continue to justify to either wife or dear daughter. Consequent belt tightening led to our Jianhan, Young Robert and Michael spending a day or three porting the existing code to Azure. All of which appears to now be working. We await the return of Librarian Jayne who’s been on vacation for the last fortnight. Great timing Jayne, really great timing. Once Jayne’s back and has declared herself happy, a migration plan will be hatched and, once that’s done, costs transferred from Michael’s purse to the public purse. Tune in next week - coughs - for more.

Thanks on this one are due to Russell who helped Robert and Michael to route round Azure’s singular inability to run a rake task. Decanting code from rake to modules and knocking up a quick script to hammer a URL every ten minutes appears to have done the job. Thanks Russell.

Legislative reform orders - some progress at last

Before packing her swimming cossie, dumping Gordon at the kennels and climbing aboard her aeroplane, Librarian Jayne managed to steer Michael to the conclusion of legislative reform order procedure mapping. Her very last job was adding step labels to the maps, at which point we waved her merrily off on her travels. Which meant she had the good fortune to neatly side-step both sitting day changes and the somewhat tedious matter of adding procedural routes to machines. A job that fell neatly into the lap of Librarian Ayesha and Michael. And a job that has so far taken them the best part of a week. As of Friday, the poor machines have been forced to ingest three component maps. Only four more to go. At which point, we need to check it all. And if Jayne thinks she’s avoiding that too, well, she has another think coming.

In bill mapping news, we’ve had some difficulty getting hold of clerkly colleagues in the Northern Ireland Assembly. Undeterred, Jayne fired up her internet browser and attempted to decipher the runes of NIA standing orders, committee reports and other assorted business papers. Brow much furrowed, she made a first attempt at a procedure map for the NIA legislative consent motion procedure. Something of a punt, but better than nothing. Then Bill Office Liam worked his magic, delivering a gentle prod in the direction of Belfast and contact has now been made. A meeting being pencilled in for next week. At which point we learn if Jayne’s reading was correct or sadly misguided.

Once the NIA map is in place and signed off, we’ll have LCM procedure maps for all three devolved legislatures. We’re around three decades away from mapping the full bill procedure, and nevertheless we’re starting to wonder if the bits we have might be put to some use. At least for tracking LCMs. Our procedure model was designed to cope with steps occurring in Houses, the matter of steps in legislatures without Houses not really crossing our minds back then. To solve that small problem, our Jianhan has been hard at work updating our procedure editor database, procedure editor application, triple store, pipes and plumbing. Which means we’re close to being both bicameral and unicameral compliant. Not something you can say about most software.

Sticking with matters of consent, but back over in secondary legislation land, Jayne and Michael have made a couple of small map changes to cope with written statements made under section 10 of the United Kingdom Internal Market Act 2020. The new steps saw their first outing on 20th July for the United Kingdom Internal Market Act 2020 (Exclusions from Market Access Principles: Single-Use Plastics) Regulations 2022.

A small tweak to treaties

Many of our procedure maps start with a step called ‘Instrument created’. It serves no real purpose and carries no real semantics, but, if you’re gonna parse a procedure, you need to start from somewhere. It turned out that, in the case of our CRaG treaty map, we’d been decorating the ‘Instrument created’ step with the date on which the treaty first came to our attention. Which basically meant the laying date. This caused some confusion amongst our International Trade Committee colleagues who pointed out - quite sensibly - that it would be more accurate to say a treaty was created on the date it was signed. But where would we get such information? Treaties are well known for having titles that stretch h1s to breaking point. A good example being the Free Trade Agreement, done at London on 16 December 2021 and Adelaide on 17 December 2021, between the United Kingdom of Great Britain and Northern Ireland and Australia. At no point had we stopped to ask treaty-minded colleagues for the definition of ‘done’ or why the thing might be ‘done’ twice. Dear reader, it would seem that ‘done’ in this context means signed and there’s two dates because there’s two time zones. This is possibly something we should have spotted. In the treaty map and the treaty map only, the ‘Instrument created’ step has now been replaced with a ‘Treaty signed’ step and all dates updated accordingly. We can only apologise for earlier stupidity. We’re still inclined to think ‘done’ is a silly word though.

On matters of modelling

Nothing ever quite being finished, we returned to our beloved procedure model. Some time ago we added means to handle citations: this procedure is informed by this Act, this route is informed by this section of this Act, this step is informed by this standing order. And so on. And so forth. And we found many more things require citations and many more things are cited. So now we’ve stripped out the citation elements from the procedure model and created a new model dedicated to all things citation-like. It should, in theory, allow us to cite both modelled things in our triple store and things way outside our control by means of a URI.

In more exciting modelling news, we’ve taken everything we learned from our chats with clerks, librarians and the TNA’s very own Helen and cobbled together a new and much improved laying model. At this point, the keen eyed clerk might be thinking, “hang on, you’ve missed all the important bits. How are we supposed to know when a paper was laid? And by whom? And in which House? This doesn’t seem very good at all. Get back to fixing printers you blithering idiots.”

In response, we note that Librarians Anya, Jayne and Ned, together with their computational helpmates Young Robert and Michael have also made a brand new and rather shiny making available model. In time, this is intended to function as a superclass for all the ways in which a thing may be made available to a House, covering laying, depositing, presenting, recording and so on and so forth. For now, we only have our laying model so it’s a superclass with one subclass. Give us time and we’ll get there.

Ned and Michael have also made a couple of tweaks to the peerage model to cover occasions when a person is accidentally summoned to the House of Lords despite not being in possession of the necessary peerage. Accidentally being slightly problematic here. It not being possible for the monarch to make a mistake, such a summons is deemed to have created the necessary peerage ‘by writ’. All of which means we end up with peerages with no accompanying Letters Patent. A situation our peerage database is not currently equipped to deal with. How we update both application and database to cope, we’re not entirely sure, having run out of any and all means to actually pay James. A problem for another day.

Away from their usual ontological duties, Young Robert and Michael have been press-ganged into a project to design data models for the Registers of Members Financial Interests. We pause to briefly practice our best Paddington Bear stares in the general direction of the Woodhams lad. So far, they’ve been working with the House of Commons registrars to turn the ruleset and accompanying Word forms into something roughly resembling a set of relational data models. Whether that’s going well, we’ll leave others to judge. They’ve also been chatting with Steve about competency questions and how to ‘socialise’ them. As Robert likes to say. On Wednesday night, Michael took a trip to Newspeak House to chat to Edward about the possibility of - and with apologies for the dreadful word - crowdsourcing some more. ERDs aside, they believe they’ve come up with something that at least roughly resembles a plan. Conversations with a clerk from the Standards Committee suggest that, whilst the rules are immutable in the short term, their compilation might not be. Which hopefully means the project becomes a set of negotiations between rule interpretation and competency questions and Robert and Michael can sit back and await the results. We’re currently in the midst of commissioning Open Innovations to build a disposable prototype, which, if things go well, should allow us to easily adapt the models and see what changes. This for both inputs and outputs. Tune in next week if RMFI is a thing that interests you.

On matters of information management

Recent events have proved a fertile testing ground for our crack team of librarians’ crack librarian skills as monarch gendered issues raised their heads almost everywhere. Team:Phil have been hard at it, updating reference data from every angle possible. Members who were Queen’s Counsellors are now King’s Counsellors. Her Majesty’s Official Opposition are now His Majesty’s Official Opposition. King’s speech debates are a thing, a new sibling term to Queen’s speech debates, under the new Monarch’s speech debates to make querying for either easier. And so on. And so forth.

A wee while back, as their remit extended and their tentacles stretched out into more systems, our librarians decided to take a new approach to manual making. Alongside their existing manuals per service, they also created cross-service ‘what to do when’ guides covering eventualities such as prorogation, dissolution, general election, death in service and the election of a new Speaker. Unfortunately no-one thought to make one covering the death of a monarch. That is now fixed. Obviously.

Probably our most exciting bit of information management news comes from Librarian Jayne who has now added all pre-1987 Acts of Parliament necessary to give all our SIs an enabling Act. This to both Search and Indexing and the data platform. Show more options to give it a whirl. Top stuff Jayne.

Shortly after the last general election, we took a feed of newly elected Members from our election database and used it to create skeleton records in our Rush database. Sadly the general election database no longer exists, so nobody call an election too soon. Anya and Michael also took the opportunity to trek down to Devon and pick up the Professor’s latest set of index cards. Taking in a public house or two on route. Index cards safely collected, events once more overtook intent and not much actually happened. This has now changed as Librarians Anya and Anna have started the long and laborious task of decanting details from index cards - it would appear that elevation from Doctor to Professor does not necessarily improve handwriting - and into the database. Mostly Anna if we’re honest. Poor Anya managed a happy day on the task before being buried once more under a tottering tower of managerial administrivia that seems to come with the territory around these parts. In the course of their work, they came across a handful of previous failed parliamentary election candidacies that had been missing from MNIS. Those are now added. Lovely.

Staying with the Rush database, our crack team of librarians had a long and interesting - to librarians at least - chat about the difference between NULL values and ‘unknowns’, eventually emerging on the far side with a brand new policy and a manual update. Meanwhile, Librarian Anna has tidied up the reference data for political organisation connection types and James has normalised the results into a new database table and given the resulting lists URLs. Because giving things HTTP URIs is just how we roll. Top work Anna, top work James.

In other team:Phil news, the work to tidy up contact detail information stored in MNIS trundles ever on. Details for ex-Members of the Commons have now been expunged - at least in cases where elevation to the Lords hasn’t happened - and what’s left has been queried and tidied. We now have an established style for telephone entries and blushing enquiries sent to current Members whose telephone number included the word ‘fax’. It’s cheaper than buying a fax machine. Phil even wrote a nifty little script to check that links to Members’ personal websites, constituency websites, social media accounts etc. are all returning 200s. Then promptly lost it. The code has been rewritten and is now on GitHub. He’ll not make that schoolboy error again.

Box fresh ‘brarian

Usually we have the pleasure of introducing a brand new data model. This week we welcome a box fresh ‘brarian to our flock. If flock is indeed the collective noun for a group of ‘brarians. Other options have been suggested, a “stack of ‘brarians” being one favoured variant. With that, we welcome Deanne as the latest addition to our crack team. Whether she’d prefer to join a flock or a stack, we leave as a matter for her. Welcome Deanne.

We also welcome back Librarian Claire from her short but refreshing career break. Lovely to have you back Librarian Claire.