It was a four-day week this week due to the May Day holiday on Monday.  On Tuesday I spent quite a bit of time responding to a couple of emails from Tony Harris about the Bilingual Thesaurus system, which he is repurposing for another project.  This involved going through the code and database for the Bilingual Thesaurus and refamiliarising myself with everything in order to reply to Tony’s questions.  Whilst doing so I also spotted that one of the lexemes in the thesaurus was classed as a noun when it should have been a verb, so after checking with the PI Louise I corrected this.

Also on Tuesday I finally managed to register for the DH2025 conference with the help of Critical Studies Admin Emma McCluskey, and it’s great to have everything sorted for the conference now.  We also tried to sign me up for the DH even that is happening in Glasgow in June but unfortunately encountered some issues with the registration process and were unable to proceed.  I’ve contacted the event organisers and hopefully I’ll be able to complete the process next week.  I also fixed a minor issue with the pilot website for the Books and Borrowing project and replied to Kirsteen McCue about a collection of songs and accompanying documents that we’ll be adding to the Burns website soon.

On Wednesday I focussed on preparing my presentation for the DH2025 conference.  I’d met with Jennifer Smith last week to discuss this and she’d prepared an initial Powerpoint presentation, and my task was to reshape this and prepare the actual talk around the slides.  The structure differs slightly from Jennifer’s original, with the biggest changes being the inclusion of a brief discussion about SCOSYA at the start, to give a bit of context for the creation of SFY, and also the inclusion of a section describing how the maps are generated as I figured the audience are likely to be more technical than linguistic and might be interested in this.  There’s quite a lot of content, but I timed myself reading the text and it came in at about 8 minutes, so hopefully it will be about right.  Obviously I’ll read through it many more times and will tweak things if needs be.  I also sent it on to Jennifer in case she would like me to make any changes or has any suggestions.

On Thursday I had an online meeting with the Dictionaries of the Scots Language to discuss our recent rollout of the new version of the DSL data and accompanying new website features such as the sparklines.  This was a very useful meeting where we all had a chance to discuss the process and any aspects that could be improved in future releases.  Also mentioned at the meeting was that the team wanted to reinstate the ‘filter results by dates’ option on our test site for further testing, which I did after the meeting.  I also shared the scripts and instructions needed for generating a new version of the DSL data from the XML files outputted from the DSL’s editing system so the team has a copy of everything in case I become unavailable for whatever reason.

The editor Ann Ferguson also spotted an issue with the bibliographical links referenced in some entries.  These were not displaying as links and were therefore not leading to the relevant bibliographical entry.  It turned out that the possibility of there being authors and no titles in the citation records had been overlooked in my code and in such cases the link to the bibliography wasn’t getting added in.  Thankfully once identified it was relatively quick to fix – no need to regenerate the Solr index or anything like that.  I don’t think the update will have introduced any unintended consequences so I therefore updated both the live and our test site to implement the fix.

On Thursday and Friday I also did some work for the Anglo-Norman Dictionary.  The editor Geert wanted me to update the new cognate reference section in the advanced search to have all cognate dictionaries unselected by default, which I did.  This does now mean that the cogref search will not submit when someone types in some text and presses ‘search’ – they will always have to make a dictionary selection first.  However, Geert seemed happy with this.

Geert also wanted to make the ‘Advanced Search’ option more prominent in the site’s navigation and suggested moving it from the site menu to a separate button near the ‘Quick Search’ section in the top right of every page.  I pointed out that we need to retain the ‘Advanced Search’ menu item in the menu bar as all major sections of the site are represented in this bar, and one is always highlighted.  If we removed ‘Advanced search’ from the menu bar we’d end up with no item highlighted whenever a user is looking at the search form or search results, which I think would be confusing.  However, I agreed that the advanced search could do with more prominence and proposed that we retain the ‘Advanced Search’ menu item, but we also include a ‘Try an advanced search’ link underneath the quick search bar in the header.  I favoured text rather than a button, as a button would be too prominent and would take the user’s eye away from the actual search button.  I sent a mockup image of this to Geert and he agreed with me, so I implemented the update, which is now live on all AND pages.

Geert also mentioned that the bibliography (https://anglo-norman.net/bibliography/) was very slow to load and I spent much of Friday reworking this.  The bibliography page was loading all 1300+ sigla and then limiting the display to those with titles matching the selected letter.  This was not a very efficient way to handle things and must have been implemented back when the bibliography was much less extensive.  Instead I updated the API and the bibliography page so that only those sigla that actually begin with the selected letter are returned and processed.  This has massively increased the speed at which the page loads.

I also noticed that the ‘Citation Siglum’ search box in the ‘Citations’ tab of the advanced search was not functioning as it should do.  As you type into this box a list of sigla that match the characters you enter should be displayed, allowing you to select the appropriate item.  However, no matter what text you entered the entire list of more than 1300 sigla was being displayed, which took time to load in and was utterly useless.  I also fixed this and now (for example) when you type in ‘roy’ you are presented with a list of sigla that actually feature these characters anywhere in their titles.  I’m not exactly sure when these issues arose, but I’m guessing it was when we reworked the sigla to greatly expand the information included (e.g. items, links to cognate dictionaries etc).  I’m glad Geert brought the issues to my attention and I was able to fix things.