Last week was all about meeting actual, living human beings in actual meatspace, taking to whiteboards, chatting, sketching, chatting some more and learning lots of new things. It was all very pleasant. Very pleasant indeed. This week was all about documenting the new things we’d learned. Before we forget them. Again. So, on the whole, much less pleasant. But at least we have something to show for our efforts.
Librarians Jayne and Claire, together with computational dilettante Michael, took to Omnigraffle and text editor to turn what we learnt from two sessions with Richard and Graeme into models and maps. Which means we now have a very basic - very, very basic - model for bills. It’s currently in a half commented state and we know we need to add a bill class and some subclassing to public bills, private bills and hybrid bills and generally jiggle things round. But it’s a start.
They also took what Richard and Graeme taught them about public bill procedure, combined it with earlier efforts and started work on a brand new logical and arithmetic procedure map. So far we’ve covered how to determine a bill has left a House - fairly easy; how to determine if a bill is in a House - a little harder; and how to determine if a bill is able to enter a House - really rather tricky. Fortunately, it was all made rather less tricky by Graeme’s revelation that a public bill carried over into a new session is in fact a different bill. Which means our map doesn’t have to take account of carry over motions or the invocation of the Parliament Act. It is rare to emerge from a modelling session with the realisation that things are actually less complicated than you imagined them to be. Long may that continue.
We must confess we are still a little confused by the difference(s?) between a first reading in the House of Commons where the House of Commons is the first House encountered and a first reading in the House of Commons where the House of Commons is the second House encountered. Graeme informs us that a bill enters the first House by means of a ‘presentation’ and enters the second House by way of a ‘message’. Which the second House may then consider. Where this leaves a Ten Minute Rule Bill, having had its motion for leave approved, we remain less than sure. Should anyone be able to explain matters further, we’d be most grateful. Martyn / Paul, it’s possible we may be looking at you here.
More documentation was generated on Thursday when Librarians Anya and Ned and computational wizards young Robert and Michael took everything that Carl and Mark had taught them about geographic areas and applied their tiny ontological chisels to a whole new model. Again it is only half commented and half of those comments are buried in a couple of containment properties. Once again we find ourselves mithering about making models that obviously express the domain in a manner that is amenable to people, as against subclassing and enumerating and so on in a fashion the machines might favour. It always feels rather unfortunate when hard won domain knowledge gets buried in comments.
Yet more documentation emerged from lasts week’s chat with Joe and Nodar on the nature of system dependencies in the answering of written questions from the House of Lords. Michael has taken what they told us and glued it atop everything Table Office Matt told us, as our swimlane style visualisation of workflows continues to evolve.
Still, it was not all pens and pencils. This week Anya and Michael also met Hansard Sally to chat through the approaches she takes and the buttons she presses as part of her role as House of Lords gatekeeper. Depending on how the rest of Saturday goes and whether or not a snooze is felt needed, by the time you read this, that chat should also have made it into our process flow diagram. Or not.
Not content with the combined brains of Sally, Matt, Joe and Nodar, this week also saw a session with Librarians Anya, Anna, Emily and Steve to draw through what they do with written questions and answers. Quite a lot of linking it would seem, whether that be linking to subject terms or linking to other related material. For now, this remains in sketches. Meaning more documentation is needed on the far side of half term.
At the behest of bossbloke Ian, Michael has also been busying himself with mocking up instance data for House of Lords questions tabled for written answer. A couple of changes to our question and answer and tabling models have dropped out of that work. At this point, it is only right to thank Silver for pointing Michael at Nick’s mighty fine Easy RDF convertor and the marvellous Mr Humfrey for making such a thing. It has saved hours of work already. Michael always knew Nick would amount to something someday.
On the subject of domain modelling and the simple joys of chatting and sketching, Anya and Michael were fortunate to be invited to give another talk, this time to the ISKO UK crowd and covering pretty much everything we know about building services from the domain model up. With the focus quite firmly fixed on why - in an organisation of moderate size and with any degree of complexity - it’s often so bloody hard. Once again, the event was virtual in nature with full use being made of Tothill Street’s new ‘teleconferencing facilities’. Which is basically a telly with a camera that would not play nicely with Zoom. Anya and Michael wish to apologise for staring blankly into space for much of the presentation but it’s important to realise the slides and notes were on the telly and the camera was on the laptop. More practice needed here we feel. Blank stares aside, it seemed to go pretty well. Though we can’t pretend we weren’t disappointed when - following 30 minutes of explaining why sketching out conversations is more important than toolsets - the first question we got was what toolsets do you use? CHATTING TO PEOPLE AND WHITEBOARDS being the only appropriate answer.
Young James has continued to make marvellous progress on taking Librarian Anna’s spring clean of the Rush database and normalising the newly tidied data into more easily managed tables. We now have lookup lists - or reference data if you will - for person titles, genders, education levels, countries and ways in which a person may be connected to a local government. Marvellous stuff.
Staying with James, he’s also put the finishing touches to the development work on the database of peerage creations we’ve been beavering away at. Librarian Ned chipped in to help test and we have now declared ourselves more than happy. Because the code side was largely based on what he’d already done for the Rush database, the development work was kindly paid for by the History of Parliament Trust and the resulting website will be hosted by them. Thanks Paul. Unfortunately, we can’t yet link you to what James has built because Ned and Michael’s testing has made a right old mess of things. But we can link you to an earlier version we built, should peerages, letters patent and people being issued with peerages be your bag of chips. We only await David supplying a new data dump with a few fixes to announcements and we should be ready to go live. Applause everyone.
And finally, an email to Parliament and Constitutional Centre Chris in search of a prioritised list of features to add to our MNIS prodder came back with a request for the inclusion of filters for Members being returned at a given election and Members representing a constituency in a given - NUTS 1 - region. We are delighted to announce that both of these filters have now been added.