It’s not every week we hand out one of our much coveted Librarian of the Week awards. Such commendations are not baubles to be thrown around hither and thither, nor indeed willy and nilly. It’s even rarer to see Librarian of the Week bestowed upon a non-librarian. This week, we make an exception, as our Jianhan becomes the first recipient to have never attended library school. This for turning Michael’s whiteboard fever dreams into actual working code, keeping Librarian Jayne sane, keeping our procedural work on the road, and almost single-handedly resuscitating an antiquated yet somehow live search service that hasn’t been upgraded for 13 years. He really is our single-point of computational success and, without him, we’d be quite, quite lost. Congratulations Jianhan. And thank you.
Decent progress continues to be made on the frontend of new old search. Designer Graeme has been hard at work churning out new wireframes for new content types, developer Jon has been busy turning Young Robert and Michael’s code scribbles into something more legible to machines, and data analyst Raafay has come aboard to investigate which attributes populate which content types in our antediluvian instance of Solr. Our Jianhan, as ever, lending a much needed hand.
With the benefit of hindsight, the data analysis work would have kicked off well before the design and development. But we didn’t clock that back then. So we - and by we, we mean Rafaay - are scrabbling to get analysis done ahead of work that’s already half happened. One lives and one learns. Anya and Michael have put together a new Trello board which they’re hoping might help, though it feels rather more like a production line with dependencies than anything parcticularly Kanban-like. We mustn’t let project manager Yomi know, or he’ll be demanding a GANT chart.
The backend work for new old search is proving trickier. So far, our Jianhan has managed to upgrade our superannuated Solr instance from version 3.5 to version 4. A spike - or safe to fail probe, as we like to say in these parts - that, rather unfortunately, saw some of our content go missing. Anya suspects she’s identified why, which is some comfort. So just another five upgrades - and a full reindexing - to go and we should be getting somewhere.
A meeting with boss computational expert Ian revealed we have more work to do to replicate both Solr and SES onto the public internet before we can begin to think about making new old parliamentary search public. We’re pretty sure making websites didn’t used to be this complicated. A meeting has been pencilled in for Monday with Young Robert, Michael and - all too obviously - our Jianhan in attendance. A document may well need to be written. We promise we will use as many normal words and as few acronyms as we can get away with.
Last time out we reported that our Jianhan had once more ploughed through Trello tickets, every box ticked on our mission to transfer control of display order for statutory instrument procedures to the more than capable hands of our crack team of librarians. But a computational expert - even one as capable as Jianhan - is nothing without a trusty librarian to rely on. So this week, we’re pleased to report that Librarian Jayne has rewritten our SI website SPARQL queries and passed them on to colleagues in the Software Engineering team. Whenever they get a spare moment, we hope to see them deployed to staging for testing, then - all things being well - into live. At which point, JO Jane may well owe Jayne and Jianhan an ice cream.
Work on planning for the next general election continues “at pace”. Back in week 29 we reported that Sym had added a new slot to the Democracy Club database to store MNIS Member IDs, using a nifty little Wikidata query to populate them. Lovely stuff. Librarian Anna has spot checked Sym’s work and, happily for all concerned, signed it off with her special librarian rubber stamp. Top work Anna. Top work Sym.
A retweet from statistician Carl alerted us to the fact that the ONS have minted identifiers for the post boundary change constituencies. At least for England and Wales. So our next job is adding those to our constituency spreadsheet. We eagerly await the same for Scotland and Northern Ireland.
In some ideal world, with the commissioning and building of any new system, feeds for downstream integration by our crack team of librarians would appear on the to do list by default. Unfortunately, we do not live in an ideal world. Time, money and attention spans all constraining capability. Such was the case with our new committee and bill systems, leading to teary-eyed librarians entering their fourth decade of combing Votes and Proceedings on a semi-domesticated goose chase for news of oral evidence sessions and bill publications. As the 35th anniversary slipped silently by, Anya explained the problem to Michael over one of their many team-building cigarette breaks and Michael returned to his laptop to see what could be cobbled together from the APIs we do publish. All of which led to new feed-bearing applications for bill papers and committees.
In the first instance, the committees feed app only dealt with oral evidence transcripts. Following a request from our reader, Michael was knee deep in adding written evidence when Anya asked him to put down his pen and concentrate efforts on committee publications instead. Which he did, returning some weeks later with a proud smile on his face and the hope of a pat on the head. Unfortunately, Anya had no recollection of the conversation and little idea why committee publications might be of interest to her librarian brethren. You just can’t get the managers these days.
Fortunately, Librarian Jayne did remember the request, which had apparently originated in a need for timely information on committee correspondence with government for inclusion in our treaty tracking service. So Michael added aggregations by committee and inquiry and Jayne was indeed happy.
Young Robert was also happy but no less demanding than he ever is. Vacationing on the continent, he telegraphed to say if we’re gonna be adding RSS and ICS and JSON everywhere, we should probably point that out to crawlers and other robots. So Michael spent another few hours adding rel-alternate autodiscovery to bill papers, committee feeds and our beloved egg timer. Well, if a job’s worth doing, it’s worth doing proper, as Robert might have said. But he was paying by the word.
Librarians Ayesha, Jayne and Claire continue to plug away converting 41 fairly intricate procedure maps from Omnigraffle to Visio. Something of a thankless task. 29 are now considered done, so the back of the job is pretty much broken. As part of that process, they’ve also been adding arithmetic components to some of the maps that predate our invention of procedural maths. 11 of 11 now being converted in both map and data. Top work librarians.
We are once again delighted to announce the publication of yet another refactored and revamped spreadsheet in our Parliament: facts and figures series. This one an effort by Librarian Jayne on the subject of expedited legislation. A big improvement. Lovely stuff.
We’re often told there are two hard things in computer science: naming things, cache invalidation and off by one errors. That’s a little computer joke for you there. It’s one we’ve used before. To be honest, it’s the only one we know.
The things needing naming in this week’s episode are ships. Amongst a great many other things, our crack team of librarians are responsible for maintaining the Parliament Thesauri. A set of synonym-rich, polyhierarchal taxonomies used to subject and procedurally index parliamentary material for the purposes of Parliamentary Search. Some years ago it was decided to add ‘scaffolding’ terms to better group taxons, one of which was ‘Names of ships’. The separation of types and instances always felt like a fudge, but it allowed our user to browse two lists - types, being boats, ferries and hovercraft; and instances, such as Exxon Valdez, Mayflower and Rainbow Warrior.
Over the years, a great many ships have been brought to Parliament’s attention. A vertiable armada. HMS Astute, HMS Bulwark, HMS Caroline being just three examples. Librarian Ned took a look and decided that ‘Names of ships’ was rapidly morphing into ‘Names of terms beginning HMS’ and the whole thing could use a dash more structure. Which means HMS Cornwall now has a broader term of Type 23 frigates which has a broader term of Frigates which has a broader term of Warships. Lovely. The old ‘Names of ships’ scaffolding term has been renamed “Names of civilian ships”. A renegotiated compromise, says Librarian Ned.
