Week 42. My word. Who’d a thowt it. As the long hot summer fades into misty mornings and mellow fruitlessness, we find our team of crack librarians and computational experts busily avoiding work in search of self-improvement.
Actual computational progress slowed this week as our Jianhan took a highly-earned, long-overdue vacation. Despite this slight setback, Librarian Jayne and computational aficionado Michael had approached the week with high hopes. This, they had thought, would be the week when a whole new procedure was translated from pixels into data. Champagne corks would pop and hair playfully ruffled as their fanbase celebrated the arrival of the draft composite procedure to the world of digitally connected devices. Sadly, their dreams came to naught as the pipes from data platform to website remained resolutely blocked. Our colleagues in Software Engineering continue to investigate the cause. A lesson for us all in disposing of parliamentary procedures in an appropriate fashion.
Nevertheless, some small progress was made. More of a gesture perhaps, but not nothing. On Monday, our super-keen, fresh-faced compatriots finally finished adding the motion to extend period A to the Commons side of both the treaty map and the data. Yet more hot mess in a sea of hot messiness. Thanks Constitutional Reform and Governance Act 2010.
And on Friday, we finally got around to dealing with the current suspension of Standing Orders setting out the rules for English Votes for English Laws on instruments of a statutory nature. Suspended, we believe, until two days before Bonfire Night. Thanks Brigid. Procedural routes suggesting some steps caused EVEL certification have now become routes allowing certfication. Jayne and Michael were about to add a new step for Standing Order suspension, but - unlike suspension of scrutiny reserve in the Lords - this is not on a per-instrument basis. So we’d have ended up with bubbles floating free of all connections. And we’re not really sure what benefits this might bring. The card has been moved back to our ‘brarian discussion column in the hope someone else might have better thoughts.
A question on Twitter from Arabella threw Wednesday into chaos and confusion. She wondered aloud what the last laying date might be for Parliament to pass - or at least not reject - an EU trade treaty before the year ends, given CRAG’s 21 day scrutiny period. Michael plugged the numbers into his trusty egg timer and watched in horror as the glass broke, the sand flew everywhere and the egg timer remained resolutely silent on the subject. A quick investigation revealed that the database didn’t have enough future dates to perform the calculation, despite our Google calendars being populated way into the future. Further investigations revealed that the Google calendar API only returns a maximum of 2,500 results, after which it demands you paginate. Michael cursed at some length that this fact had not been in the documentation. Then he used Command-F on the page and found it. Which should be a lesson to computational experts everywhere. Michael and Robert dived back into the sync script, added pagination through result sets, and retired bruised but happy.
The answer, Arabella, should you still be interested, is the 17th November. Laying a day later pushes us out four days into the new year. Although, the 17th means both Houses sitting on Christmas Eve. Which seems a lot to ask.
In the course of his flailing attempts to debug the sync script, Michael discovered his attempts to render a full session’s worth of dates, together with a touch of introspection about how Parliament might describe those dates, was fine in theory but poor in practice. For long sessions - of the kind we’ve grown used to - so many queries were running that the request timed out. So the rest of Wednesday was spent removing lists by session and adding lists by month. Which our Robert stepped in to prettify a little.
On Wednesday, librarians Jayne and Anya, together with computational crackerjack Michael, had another meeting to plan how our librarians might tackle the task of taking care of our calendars. The previous plan of using recurring events had proved confusing for everyone. Not least Michael. So they decided firmly against continuing that course of action. From now on, we’ll aim to populate 60 days in advance and review where we are every Monday. Whilst still reacting to announcements from the Houses as expeditiously as possible.
Efforts continue to resurrect the data platform to something like functional form. On Thursday, Anya, Robert and Michael used up more pixels to meet with Silver and Julian from family friends Data Language. Data Language are looking into what went wrong with orchestration of Member details from the Members’ Names Information Service to the platform. We suspect that some problem mapping the names of their Lordships resulted in what was an automated process becoming a little more manual. And that part of the manual process involved machines sending emails to humans. And that the recipients of those emails have long since left the building. Suspicions were confirmed in part by our dear but departed Wojciech answering an enquiry well outside his working hours. And further confirmed by some email archaeology performed by Ben on Friday. Numerous examples of machines forlornly attempting to communicate with humans over the medium of SMTP turned up, alongside vague memories of a spreadsheet and possibly an Access database. That last bit does seem unlikely. So we’re two steps closer to knowing what went wrong and one step closer to at least patching over it. Which does of course raise the wider question of why the mappings didn’t map in the first place.
On Friday, Ian and Michael from the Parliamentary Computational Section met with Anya and Bryn from the world of librarians to chat about some short term work on the publication of Research Briefings and some longer term work on what that service might one day look like. At least with fair winds assured. Removing the dependency on data.parliament looks like the quickest, erm, win. With apologies for the business language here. In the longer term, bringing together assorted forms of output, described in a consistent fashion and layered over with a proper model for editorial policy is a goal worth fighting for.
Robert has continued to tweak and improve his glitching of legislation.gov.uk, concentrating recent efforts on Statutory Instruments. Here, for example, are The Health Protection (Coronavirus, Local COVID-19 Alert Level) (High) (England) (Amendment) Regulations 2020 - complete with working footnotes and extremely useful, if not yet pretty, anchor link flags. Lovely.
At some point, Ian requested user stories for future management of the Parliament Thesauri. So on Wednesday, Librarian Liz, Anya and Michael took to Teams and the usual Google doc to sketch out some bullet points. They want to get to the position where the shape of the taxonomy and its interlinking with external data sources can all be managed by team:Liz without the regular intervention of our friends in DevOps. They also want to get to place where they can offer different slices and dices of the taxonomy to different consumers whilst still maintaining as an integrated whole. Or librarianship as a service, if you will.
And finally we arrive at the meat of our week. Learning. Tons of learning. Endless learning. Monday kicked off with Anya and Michael heading along to an academic panel on the prorogation of Parliament that never was. And its implications for the future. All arranged by Graeme. Thanks Graeme.
Tuesday, of course, was Map Camp. Which was mainly marvellous. And free. We were particularly struck by James’ point that the constant push toward commodity is only ever going to turn into a battle against entropy. Which is a battle you’re probably not going to win. All the presentations are now available online. And still free. Get stuck in. Thanks for organising Simon.
We ducked out of Map Camp for an hour to pop along to the latest in our long line of Lords Procedural Seminars. This time an insight into the world of divisions, presented by another friendly face. Thanks Matt.
Finally, with Covid delaying Taxonomy Bootcamp til March, Wednesday afternoon saw the first ever Taxonomy Bootcamp webinar. Librarians Anya, Liz, Emily, Steve and Ian all popped along. The emphasis on maintenance and long-term custodianship - rather than just building more random bits of software was more than welcome. Thanks for organising Helen.