On Wednesday, our own team of crack librarians met with a second team of crack librarians, also from the House of Commons Library. That second team do sterling work compiling current awareness bulletins - or what is going on in the world in assorted area of expertise - for our crack team of researchers. All of which is managed by means of an RSS aggregation service. Which works just fine for lots of externally published material. More publishers provide RSS feeds than most of us might imagine; even if those feeds are not always well advertised. Problems arise when trying to aggregate material published by Parliament, where our own websites don’t offer much in the way of alerts or content syndication.
Happily, Young Robert and Michael are of an age where publishing RSS is just part of what one does. Over the years, they’ve cobbled together a variety of prototype applications, from our Procedure Browsable Space™ to our Bill Papers™ application, all churning out RSS feeds alongside what what Robert would undoubtedly call other applicable representations of information resources. Those prototypes are aimed squarely at making life easier for team:Anya. And they’re now finding a second life of usefulness in the hands of the current awareness librarians. If you’ve used our feeds and found them helpful, why not get in touch and let us know? We’re always grateful for a pat on the head or a scratch behind the ears.
Even more happily, we’ve been putting a lot of work into new, old Parliamentary Search, allowing users both inside and outside Parliament to search across Parliament-published material using a myriad of facets and filters. If you’ve used new, old Parliamentary Search, you’ll already be aware that its default - and, for now, only - result list ordering is latest first, with each result coming complete with a title, a description, a date and a link. At which point, our dear reader will probably be thinking, what better candidate for the old RSS treatment? You would not be alone. Librarian Anya asked much the same question as she pondered how long it might take to add RSS views of search result listings. About two hours, said Michael, promptly throwing poor Developer Jon under the delivery bus.
Two days later - Jon has much higher standards than Michael, and likes to ensure everything has been properly thought through and has regression tests in place - new, old Parliamentary Search comes complete with RSS feeds. Until they’ve been tested by our own crack team of librarians and tested by the current awareness team of crack librarians, we are not yet advertising them from the HTML views. If you want to try them out, just run a search, pop ‘.rss’ between the ‘/search’ and the ‘?’ in the URL bar and you too can have all the RSS for all the parliamentary material you’re ever likely to need.
Anya also notes that occasionally - particularly during recess and prorogation periods - her own team are not always aware of new material arriving down the pipes in need of subject indexing, procedural indexing and interlinking. This is because the current ingest engine lacks anything in the way of what Robert might call ‘observability’. Jon’s new RSS feeds may also prove useful for her own crack team of librarians. Simply subscribe to a search for a given type of material, and get alerted every time a new one drops. A fine example of dogfooding, as Robert might say.
For responding so readily to customer needs and doing so in his usual professional manner, this week Developer Jon is the lucky recipient of our much coveted Librarian of the Week award. It’s overdue now we think about it, given this is his second stint with us. Choosing to return to the fold having worked with us once and experienced the chaos; not many would have been so brave. We have the budget to go large - have two awards, Jon.
Our dear reader will hopefully forgive us for continuing this section with yet another paragraph or two, but this is exactly the kind of work that always proves most pleasing. To come across a colleague with a long-standing problem that you’re able to help solve with the very minimum of effort. Not because you’re cutting corners, but because you’ve already done most of the hard work to make the easy stuff possible. And because, a good couple of decades ago, a handful of right-thinking people took the time and trouble to standardise an approach to content syndication that worked with the weave of the web. Lovely stuff.
It should be pointed out that these very notes also come with their own RSS feed, so if you’re ever left thinking, “I’ve not seen one of Anya and Michael’s massive walls of text in quite a while”, why not treat yourself and subscribe.
Our dear reader - being, we believe, of a similar age demographic and sharing similar predilections to Robert and Michael - will, by this point, have popped open their feed reader, plugged in the RSS, and wondered what on earth we’re blabbing on about. This isn’t RSS, you’re probably thinking. It doesn’t work, you’d probably add. Dear reader, it works perfectly. Just not for you. Yet.
It turned out that Jon’s RSS feeds worked fine on his machine. They also worked fine on our staging server. They would, in theory, work fine on live, safely served from the Parliament domain. That is if the CDN settings on that particular sub-domain allowed for RSS. Which it turns out they don’t. It’s now seven days since these notes made it to draft status. Requests have been made. Tickets filed. But I think we’d all recognise that the processes somehow take longer than the work.
In other new, old Parliamentary Search news, Developer Jon remains busy fettling aliases, two stacks signed off by the librarians and ready to go live. He’s also fixed a small bug arising from the case sensitivity of aliases, Librarian Jayne noticing that production search did not care a fig for casing but Jon’s code does. Or at least did. It now works for both Datecertified and datecertified. Or will do when the alias code is released. Which we’re hoping will be quite soon.
Librarian Phil also noticed a discrepancy between what production search defines as a week and what Jon’s search defines as a week. The old code insisting that a week runs from Sunday to Saturday, Jon’s code preferring a Monday to Sunday definition. This card hit the weeknotes pile with no work necessary, it being decided that Monday to Sunday was correct in all eyes except Michael’s. If it’s good enough for the International Standards Organisation, it’s quite good enough for us.
Another card requiring no effort arrived in the form of yet another discrepancy. It turns out that when attempting to return results for items with a Boolean set to false, old, old search also managed to include items with the same Boolean having a value of NULL. So not exactly FALSE, more NOT TRUE. Another thing that Jon has managed to fix in new, old search without really trying.
If our dear reader can remember as far back as season 27 of these notes, they may well remember that Delivery Manager Lydia was dragging our two resident geriatrics along on a Wardley mapping exercise for our procedural data integration systems. The resulting maps have proved quite useful, particularly in paper format. Every planning day we take our pixels down to the Library’s plotter printer, print them up at a size even Robert and Michael’s aging eyes can see, plaster them to the walls, and scribble all over them.
Which is all fine and all good, but at least some of us work in a ‘data team’ and data teams tend to be less accepting of big bits of paper than we might be. With that in mind we have collated our pixel-based nodes and our pixel-based edges into data, and with much needed help from Data Scientist Eva and Data Scientist Catherine, popped all of that data into a shiny new Shiny dashboard. Which is almost production ready, and will give us the ability to click on any given node, follow the edges upwards and build a picture of every system that we’d lose should we happen to lose the one that we clicked on. Impressive if slightly terrifying.
As if that pause for thought were not enough, we’re here to offer one more. Whilst we have no problems with Wardley maps per se, we’re not fans of some of the rhetoric around them. There is a belief amongst some circles that everything eventually tends toward commodity. A belief undermined by an old story from Mr Sheridan that we repeat often. At some point in recent history, there was a certain brand of tape head - used by our friends at The National Archives - that was indeed an off-the-shelf commodity component. At least until the Fukushima earthquake and subsequent nuclear disaster took the one factory manufacturing said tape heads offline. At which point, the kind of thing one could have almost ordered at Tandy became rarer than gold. Or indeed gnat’s teeth. So however good your map making skills, if you glance at the bit down the bottom and think that’s just commodity, we should probably outsource it, there’s every chance you might be proved wrong.
There is something about Wardley maps that brings out the nostalgist in us. They tend to point to a stage in history when globalisation was well established and seemingly under little threat. Also a stage when the fragility of the modern web was less apparent to all but the geekiest amongst us. Seen from the point of view of 2026, “why not just use some other people’s computers and move to the cloud” seems more than a little naive. But then such thoughts are well above our particular pay grades - and this is weeknotes, not Gartner - so we’ll park those reckons. If anyone out there has given any consideration to what a value chain map might look like in a world that appears to be retreating from globalisation, why not get in touch? Very occasionally, people do actually do that, and we’re always delighted.
For reasons that escape us, somehow or other we invited feedback on our Procedure Browsable Space™ from ‘senior stakeholders’. Luckily Mr Korris is a nice man and we always aim to please. For that reason, last time out we were delighted to announce the arrival of our ‘secondary legislation landing page’. Since then, Jayne and Michael have delivered exactly the same thing, but for treaties. As we type these notes, we notice that more feedback has flooded in from Mr Korris. But also, it is Friday afternoon. And nobody sane would open an email containing ‘senior stakeholder feedback’ on a Friday afternoon. Rest assured, we will get back next week, Matt.
In new query news, Librarian Jayne has adapted the SPARQL used to return a list of work packages for a work packageable thing to also include the laying body. Which is a nice improvement, but does mean we have to do the same for every other work package listing. We do know how to make work for ourselves.
Following a meeting with Librarian Anya, Jayne has also updated the query responsible for slicing and dicing the list of documents for a given work package to include a new step collection or two. Still work to do in that regard, but we’re getting there.
And finally, in terms of query refactoring, Jayne has worked her way through every query responsible for rendering House pages and every query responsible for enabling legislation pages - including our lovely little legislation.gov.uk lookup tool - all of which come complete with comments and a link to our SPARQL endpoint. So if you’re not happy with our efforts, why not roll your own.
In the background, our Jianhan chipped in with his usual helping hand, which means a stray step depth is finally deleted and a couple of duplicate government departments are nixed. Thanks Jianhan.
Also in the background, Librarian Ayesha has been on cartographic duties, removing any and all mentions of motions not being called from each and every map and their accompanying data. That work amounting to a grand total of 27 maps free of ‘not calling’ detritus. A herculean effort, as we have come to expect from our beloved brarians.
Over with team:Phil, there was the small matter of yet another by-election to deal with. And dealt with it was. A tag team of Librarians Deanne, Harry and Phil stepping up with the usual applomb. Following on from that, the Liberal Democrats went with a reshuffle of their spokespeople. Another update issued with no hitch. We’re getting quite good at this.
Finally, a search of our intranet pages for MNIS used to yield a page mostly describing the existence of a database with member data in it. And a search for members website returned results but little relating to the website with the members’ pages. Given most of our members have no clue what that database is, or why updating it might reap any reward, we’ve added a new page more focussed on how to update their details on the website. Skipping over the somewhat uninteresting details of how the computers happen to deal with such matters.
Team:thesaurus continue to deal with feedback from Commons Library researchers, based on their initial exposure to our Library Knowledge Base™ in its current Subject Specialist Finder™ guise. The past few weeks have seen updates to labels, alternative labels, scope notes and taxonomic position of concepts as far and wide as Court orders, family law, famine, Gurkhas, medical equipment and natural diasters. That team have also issued a whole slew of changes pertaining to matters in and around climate science, net zero and climate change adaptation.
Given that the Library Knowledge Base is the first service that’s ever exposed the taxonomic structure of the thesaurus to end users, Librarian Ned has also proposed a new policy on what our crack team of librarians refer to as ‘scaffolding terms’. Or concepts in the thesaurus that are not intended for use by indexers, but exist purely as managerial holding pens for related concepts that are. That policy proposal being met by applause from across our librarian community.
Emma and Jayshree have completed the herculean task of revisiting the indexing of 6,914 briefings, from 2014 to 2025, from all three of Parliament’s research services. An oddity of briefings is that they’re indexed twice and using two separate taxonomies; once by authors selecting terms from a “topics” taxonomy of 300 terms, for browsing purposes. And once by librarians using our beloved thesaurus, for searching purposes. Old, old search doesn’t distinguish between the two kinds of indexing, and so a decision was made - long ago, when old search was new search - to not repeat concepts in the more granular indexing if they had also been used for the browsing indexing. Like we had some idea of how relevancy works. Our Library Knowledge Base uses only the thesaurus terms, and so some briefings were certain to fall through the cracks. Thanks to Emma and Jayshree’s tender ministrations these briefings are back where they belong. And thanks to new Search being in the wings, that old indexing policy is safely in the bin.
Outside of Library Knowledge Base™ usage, there has also been an update to the planning concept and - where disambiguation is required - entities associated with the devolved legislatures and governments now use the full name of the legislature or government and not the rather confusing acronyms we had been using.
Back on data tidying duties, we are pleased to report that all records that should be indexed with the closure motion concept have been freshly indexed with the closure motion concept. Librarian Emma picking up that task. Not content with tidying our own data, Emma has also been on Wikidata duty, ensuring that all Acts of Parliament in that world come complete with their UK Parliament identifiers. If there’s one thing you can say about our librarians, they truly are native to the web of data. As some chap once said.
Martin meanwhile has been exercising his librarian chops on affirmative instruments approved by the House of Commons. The Journal Office normally add the date of approval to their systems, which then flows through the pipes and into our systems. But for reasons unknown, 140 had slipped through the net with no approval date applied. Martin hooked some bait to his rod and went off on a fishing trip. 140 affirmative instruments were landed, all of which have their House of Commons approval date in place. Excellent catch, Librarian Martin.
Librarian Emily has been concentrating her tidying efforts on all of the places where we’d somehow used Henry VIII clauses as a procedural term, despite Henry VIII clauses not being a procedure. These have also been tidied.
Not to be left out, Librarian Ayesha has been decluttering our motion records, sorting those that lapsed from those approved. Neat.
On Wednesday of this week, ‘brarian Jayne, ‘brarian Phil, boss ‘brarian Anya, boss boss ‘brarian Bryn and computational pixel-pusher Michael skipped past the Two Chairmen, paraded down Old Queen Street and found themselves in the salubrious surrounds of the Institution of Civil Engineers. There to lend a supporting hand to Data Scientist Louie, who’d somehow volunteered for Data Bites talking duties. And a very good job he did too. If you’ve ever wondered how the House of Commons prepares for a general election, collects the data in and around a general election, makes use of that data internally, and publishes that data to the public, it’s well worth a watch.
And while you’re there, do stick around for Steve’s talk on The National Archives’ rather fine database of file formats that you’ve probably never heard of. If you think that a database of file formats cannot be made to sound interesting, Steve is here to enthusiastically prove you quite wrong. Splendid stuff.