Librarian Jayne has had another SPARQLing week. A solicitor with LexisNexis contacted Journal Office Peter to ask if the Statutory Instrument service could be used to identify EU Exit made affirmatives that had not been approved by Parliament. Unfortunately, such a filter does not yet exist. Fortunately our public SPARQL endpoint does. The folks at LexisNexis had been doing the work manually, but Jayne wrangled a query making the whole task considerably less onerous. The query returns all made affirmatives with ‘EU Exit’ in the title that had the ‘instrument stops being law’ step actualised. Applause Jayne.
And there’s more SPARQL news: whilst tinkering around with queries to help quality check our data, Jayne found that back in October, the Local Loans (Increase of Limit) Order 2019 had been debated without undergoing EVEL certification. She contacted the Office of the Speaker’s Counsel who confirmed that it would have been ‘not certified’ anyway. At the request of the Office, Jayne has put together a query which she’ll run regularly to ensure nothing slips through this same gap in the future. The query returns SIs that have not been withdrawn, have not been considered for EVEL, and either have or have yet to be scheduled for debate.
The team think that all this goes to prove - should proof be needed - that the work that’s gone into the data platform and procedure modelling has the potential to make a great impact beyond that of publishing a website or opening up parliamentary data. It’s the first time that procedural people in Parliament have had this information gathered and managed in one place. And certainly the first time it’s been queryable. As Anya and Michael would say, if Erskine May had had a computer, he may well have built a data platform and a precedence engine out of queries.
And the potential extends outside Parliament’s walls. Or ‘off estate’ as it’s known in these parts. Not only is the SPARQL endpoint public, Jayne continues to publish queries to our GitHub pages. Anyone with an interest can take a query, tinker with it and run it. Do get in touch if there’s a query you’d like to see and need some help writing.
Anya, Michael and Robert began the week plunging heads first into public bill procedure. David and Paul Evans joined them poolside for another session on modelling amendment and stand part consideration. For those unfamiliar with the people of Parliament, this amounts to some considerable procedural brain. Clearly we’re not including our brains in this calculation. It’s roughly equivalent to having Hemingway mark your writing. Their guests gently pointed out where the model has sprung a leak or two. The main one being that they’d never clocked that proposing happens in between the moving and the putting of a question. And, contrary to what seems to be popular opinion, it is at the proposing step and not the moving step that the amendment belongs to the House. This threw out most of their modelling around withdrawal of amendments and the begging of leave to do so. The newly introduced proposing step also messed up the flow from amendment consideration to amendment of amendment consideration, so this bit also needs a thorough redraw. Michael mentioned that he understood that Members first consider all amendments, and then move on to stand part consideration. Paul wondered aloud at someone being so bloody foolish as to arrive at such a conclusion. So that part was also revised. They also cleared up the question of whether amendments lapse at the end of a session. It turns out they do, so those steps have been added to both amendments and amendments to amendments.
Bits of Wednesday were spent peering at a rather messy whiteboard and transferring the scribbles to the flowchart. What’s there is intended to be a fairly generic model of amendment and stand part consideration that can be used across the procedures but, because recent modelling has been mostly dealing with Committee Stage, it’s slightly biased toward that at entry and exit points. New entry points have been added, but it turns out the Commons accept amendments from the moment the Second Reading motion is agreed and the Lords usually do not. For the time being, there are some green arrows to denote Commons things. Unfortunately, the flowcharts already use red to mark when steps are self-precluding, so for the purposes of this diagram the Lords is temporarily cyan. Sorry, House of Lords. The amendment flowchart still needs double checking but hopefully it’s getting closer. When we’re happy - or happy enough - we’ll move it across the public bill flowchart.
In modelling news, Anya, Jayne and Michael spent Thursday morning putting together their simplest model yet. The description ontology connects something that’s described by something to the thing that describes it. The intended use case is to make links between procedures or procedure routes and the Standing Orders or items of legislation that describe them. Because one always cites sources. But there’s a suspicion that there may be future uses, so it’s been kept as generic as possible.
If you’ve been following along from home, you’ll know that Robert and Michael spent some time back in week 8 poring over ELI documentation in an effort to work out how it mapped to our models. ELI has started on work to describe draft legislation but Robert and Michael have concerns that our models don’t quite align. They also had questions about how LDAPP and Akoma Ntoso plan to map to other outputs including, but not restricted to, HTML. So Michael mailed Catherine, Matt and John in search of enlightenment. Catherine kindly called by Tothill Street on Friday and, together with Robert and Michael, filled a whiteboard with LDAPP, Akoma Ntoso, CLML, HTML and the assorted bits of XSLT that make all the transforms work. Other acronyms may have been used. It was generally agreed that minimising divergence between the XSLT that TNA use to generate legislation pages and the XSLT that Parliament might use to generate draft legislation pages would be a good thing. In the name of consistency and not inventing the same wheel twice, Catherine said she’d bring this up at the next LDAPP meeting. They also had a chat about where ELI was at and some of the problems it faced. It turned out that no-one from Parliament attends the ELI meetups. Which happen to usually take place in Switzerland. Robert and Michael took note. Switzerland not being a contributory factor here.
Regular and close readers may recall that week 7 saw a diversion from the usual modelling of parliamentary procedure. Adam from the FSA brought along two domain experts to model the pig slaughtering registration process. It would be remiss to close these notes without mentioning that Adam has since published a post complimenting both Silver and Michael on their generosity and expertise. And he never once used the word shysters. So that’s a step forward.