I participated in the UCU strike action for all of last week and the first three days of this week, meaning I only worked on Thursday and Friday this week. I spent some of Thursday going through emails that had accumulated and tackled a few items on my ‘to do’ list. I managed to fix a couple of old websites that had lost a bit of functionality due to connecting to a remote server that had stopped accepting connections. These were the two ‘Emblems’ websites that I created about 15 years ago (http://emblems.arts.gla.ac.uk/french/ and http://emblems.arts.gla.ac.uk/alciato/) and the emblems they contain are categorised using the Iconclass classification system for art and iconography. The Iconclass terms applied to each emblem, and all associated Iconclass search functionality are stored on a server in the Netherlands, with the server at Glasgow connecting to this in order to execute an Iconclass search and display any matching results. Unfortunately the configuration of this remote server had changed and no requests from Glasgow were getting through. Thankfully Etienne Posthumus, who helped set up the system all those years ago and is thankfully still looking after the service in the Netherlands was able to suggest an alternative means of connecting, and with the update in place the site were restored to their original level of functionality.
I also did a bit of work for the DSL. Firstly I continued an ongoing discussion with Ann Fergusson about updates to the data. Whilst working on a new order for the ‘browse’ feature I had noticed a small selection of entries that didn’t have any data in their ‘headword’ column, despite having headwords in the XML entries. Ann had investigated this and suggested it might be caused by non-alphanumeric characters in the headword, but after I’d investigated this doesn’t seem to tell the whole story. It’s a very strange situation. The headwords are only missing in the data I processed from the XML files from the work in progress server (i.e. the V3 API) – they’re present in the data from the original API. Apostrophes can cause issues when inserting data into a database, but having looked through my script I can confirm that it uses an insert method that can process apostrophes successfully. Indeed, there are some 439 DOST and SND entries that contain apostrophes that have been successfully inserted. Plus the script also successfully inserted the headwords for entries such as ‘Pedlar’s Drouth’ (an entry with a blank headword) into the separate ‘forms’ table during upload, but then didn’t add the headword field containing the same data. It’s all very strange. And there’s no reason why other special characters or punctuation shouldn’t have been inserted. Plus some entries that are missing headwords don’t have special characters or punctuation, such as ‘GEORGEMAS FAIR or MARKET’. I didn’t manage to figure out why the headwords for these entries were missing, but I added them to the database, and I think I’ll just need to watch out for these entries when I process the new data when it’s ready.
My second DSL task was getting some information to Rhona about the Scots School Dictionary app. I sent her a copy of all of the sounds files contained within the app, wrote a query to list all entries that contained sound files and a tally of the number of sound files each of these has, and gave her some information about how the current version of the app stores and uses its sound files.
I also responded to Alasdair Whyte from Celtic and Gaelic who has a research fellowship to explore the place-names of Mull and is wanting to make use of the place-name system I initially created for the Berwickshire place-names project. Hopefully we’ll be able to arrange for this to happen.
On Friday I met with Gerry McKeever to discuss the interactive map I’m going to create for his ‘Regional Romanticism’ project. Gerry had sent me some sample data, consisting of about 40 entries with latitudes, longitudes, titles and text. I created an initial interactive map based on this data using the StoryMap service (https://storymap.knightlab.com/). I showed this to Gerry, but he thought it wasn’t quite flexible enough as the user is not able to control the zoom level of the map, plus he wanted a greater amount freedom to style the connecting lines as well – adding in directional arrows, for example. We decided that we would see if the maps people at NLS would let us use one of their geocoded historical maps as a base map, and that I would then create my own bespoke interface based on the ‘stories’ I’d created for the SCOSYA project. Gerry contacted the NLS people and hopefully I’ll be able to proceed with things once I have the maps.
I spent the rest of Friday continuing to rework the Digital Humanities Network website, working on a new content management system, completing work on the new underlying database, migrating data over and creating facilities to create a new project record. There’s still a lot to be done here, not just from a technical point of view but also deciding what projects should continue to be featured, and I’ll continue to work on this over the next few weeks as time allows.