Tobias Malmsheimer, our “IT guy” – as he liked to call himself – left the project in September last year to start a new full-time position at the Faculty of Computer Science at the Mannheim University of Applied Sciences. Tobi had been a core team member throughout both project phases, and was responsible for everything IT related in the project, as well as some library and information science tasks that had heavier IT components. His commitment to the project and the team is evidenced by his ongoing presence and help, taking out leave days to help support us on-site at the first JVMG Lab, for example, and being available to help us fix bugs and support the smooth transition of key IT infrastructure when needed.
Beyond the invisible, yet absolutely essential elements of keeping our tech-stack, servers and backups smoothly running, he also made quite a few key contributions to the look and feel of the project both on the backend and frontend side. One of his very first contributions was retiring the original Pubby frontend (developed in Java for the DM2E (Digitised Manuscripts to Europeana) research project) we had been using to enable the browsing of our data, and creating a new Pubby version 2 written in Python and building on Django. This is now the RDF data visualization interface users of the JVMG knowledge graph interact with when browsing our data.
He also worked on further optimizing our tech-stack, for example, by conducting query performance measurements for different triplestores, as well as developing and refining our indexing workflow that underlies our frontend search to the point that it now no longer relies on Elasticsearch, but rather builds on Fuseki‘s internal indexing features only.
His last major contribution and crowning achievement has been the development of the JVMG Sparql Builder, which allows users to conduct complex queries via our web frontend by interacting with a graphical user interface. Building queries this way requires no understanding of the data ontologies or the SPARQL query language enabling the process in the background.
Tobi has also been active in supporting our work on a large number of Tiny Use Cases, data quality assessment exercises, and most everything that had some major data or IT component to it. He was always ready to help build queries, sample and pre- or post-process data, debug weird or scrambled instances of encoding, hunt for elusive missing records or links, optimize code and workflows, and the list just goes on. He was never one to take center stage, but his constant reliable presence in the background made for a very stable work environment. Not having to worry about server security, firewall setups, regular backups and data updates was a boon for the other team members, and we could always be sure that whatever riddle the IT gods might throw at us (of which there were many), we will have someone to unravel the knot and provide a solution.
We are very happy that Tobi has already found an excellent permanent position, and even more so that we are still in touch on a regular basis, and we sincerely hope that the future will bring new opportunities for us to work more closely together again. Until then, we wish him all the best in his career going forward and look forward to our next meeting in-person or online!