2020 - Week 30

Procedure mapping

If you were following along from home last week, you’ll know that Librarian Jayne and her tame computational expert Michael have been dipping fingers into another one of those tricky and troublingly edge-case made affirmatives. In this case, the snappily titled Statutory Instruments laid under paragraph 5(3) of Schedule 7 to the Local Government Finance Act 1998. They created a new play-pen map and passed it around for comment. Being financial in nature, such instruments are laid before the House of Commons only. Which meant Journal Office Jane - their usual source of truth and knowledge - could not help here. Then again, if you have a Commons question who better to ask than Mr Evans? So Jayne mailed Paul and this week Paul replied to say he thought their finger painting looked, “broadly OK.” Which is pretty high praise. A pair of “Broadly OK” badges have been placed on order to celebrate this moment. Before they arrive, however, there is work to be done. Jayne and Michael need to check in with JO Jane and make sure she’s happy to add to the data. They’re confident the broadly OK assessment will count for a lot here. And if Jane is happy, they’ll pull apart the play-pen map, add the bits to the main made affirmative map, add the new steps and routes to the data and reindex everything that needs reindexing. And at that point a Trello card first created in February 2019, under the tutelage of Journal Office Jack, can finally move to the done pile.

In other correspondence news, Jayne had contacted Alex with a couple of questions about treaty procedure under the Constitutional Reform and Governance Act 2010 (CRAG) that had cropped up in our semi-regular “For ‘brarian discussion” meetings. The first concerned government statements to the effect that Section 20 does not apply. Which might be described as a means for the government to concede a procedural goal, declare that the ball belongs to them, pick it up and head off home. Alex confirmed that in the two years he’s been looking at such agreements, he’s never seen the ball disappear from the parliamentary pitch in such a fashion. In practice, where it’s not possible to meet the CRAG timetable, the government has proposed either:

Alex went on to add that, looking at the wording of the Act, he could see nothing that might preclude the Minister from laying the treaty and at any point thereafter picking up the procedural ball and heading home for tea. At least so long as neither the Commons nor the Lords have already resolved that the treaty should not be ratified. Which means, we think, we need to add preclusion routes to both the treaty map and data. The statement will need to preclude a bunch of procedural steps, and either House resolving against the instrument will need to preclude the statement. Given the map is already somewhat hard on the retinas, please wish our eyeballs well in this endeavour.

The second question to Alex was on the subject of Commons Select Committees and whether a Committee might ever report that a treaty - in their view - ought to be ratified. Again, this is not a thing that has happened within Alex’s working memory but he saw no reason why it shouldn’t. He also pointed out that some treaties might be considered by Joint Committees. Which is not a thing that our maps or data currently make possible. So a few more changes to make next week. Thank you Alex.

Jayne and Michael spent a little more time with pencils out, sketching what they could of the procedure for Legislative Reform Orders. This is a flavour of super affirmative, but in a slightly more complicated style than is usual. Rather than a draft instrument being laid, considered and allocated to one of three procedures and then being relaid subject to that procedure, the LRO procedure has initial committee consideration, negative procedure, affirmative procedure and super affirmative procedure all cobbled together on one mind-blowing map. It’s also not terribly well documented. Or at least Jayne and Michael have not found much of use so far. They’ve resorted to reading the legislation, which is not something their educations have really prepared them for. Like children reading Shakespeare, they recognise all the words but don’t necessarily understand the sentences. Let alone the sub-clauses. They’re fairly confident they know enough to improvise what happens after the switch to negative, affirmative or super affirmative. But they need help understanding what happens before that split. An email has been dispatched to JO Jane seeking someone somewhere with expertise on LRO procedure. Until the cavalry arrives, they think they’ve done what they can.

On the subject of mind blowing maps and help we are still in search of feedback for our composite procedure. If you’ve ever clerked a census instrument through Parliament, please step forward.

Your weekly egg timer update

With Robert returning from “holiday” next week, Michael hopes to spend a few hours continuing to comment the code. Hopefully to the level where a clerk can pick it up and tell us what’s wrong. Jayne has contacted the Commons Journal Office to let them know of our efforts in this area. With the Commons currently using dates that take into account sitting days in both Houses, we’re hopeful we’ll be able to make a small improvement in this area. Should anyone in either Journal Office feel they’d like an introduction to egg timing or a walk through of the calculation code, please do get in touch.

All about the collaboration

Jayne and Oxford Tom have been busying themselves with new maps allowing us to cite ParlRules Standing Orders from our procedural routes. If you squint closely at their new draft negative and draft affirmative maps you’ll see tiny blue and yellow dots. They represent how our routes are either informed by or occasionally governed by House of Commons public Standing Orders. The choice of ‘informed’ and ‘governed’ here is still very much up for discussion. They also continue to fill in the accompanying spreadsheet which is much more use if you’re interested in linking data and not just decorating pixels.

Still on the subject of ParlRules and linking up Standing Orders, Robert and Michael had a Monday meeting with Rad, Tom, Niels and Michal to chat about data and code licencing, the revised shape of the data and future plans for an API. Robert and Michael suggested nesting resources as the more usual design approach to a fairly hierarchical data structure. But either way, the API looks to be perfectly useable. Although it could probably live with a little more documentation.

The Monday call was followed by an actual day trip when Anya, Robert and Michael donned masks, escaped homes and took a train to Oxford town. With most of the University abandoned to ‘rona induced social distancing, they were aiming to find a patch of grass in the University Parks. Luckily the trains were on time, their sense of direction did not suffer any of the usual mishaps, few pubs were open to distract them, the sun was shining and they managed to arrive in time to meet Rad and Tom for a chat and an M&S picnic. More plans were made around data licencing, knowledge transfer, API design and the possibilities for future collaboration before our team took off for a walk along the canal. Which came to a juddering halt when an open pub was spotted across the waterway. More trips are eagerly anticipated.

On Tuesday, Robert and Michael were joined by Felisia and Ganesh from the Gubmint Computation Section for a chat with Adam about rules as code, machine learning, lexical parsing and the importance of well managed data to all of this. The persistence of faith that magic sand and enough Python will solve almost any problem, including organisations continuing to avoid doing the hard work to manage the information the machines need to learn from, seems to be a theme at the moment. A proposal for an external supervisory group to inform the GDS work was voiced. Hopefully providing another avenue for elders and betters to hear that machine learning builds on information management and is not instead of it. Hire more librarians we hear ourselves shout. Perhaps unsurprisingly.

On Wednesday, Anya and Michael reprised their What would Erskine May do? talk, first outed at the annual Study of Parliament Group conference back in January. Which starts to feel like years ago. This time they talked through their work on procedure modelling to a bunch of data interested people in the House of Commons Library. It seemed to go quite well. Though heckling is harder with microphones muted.