CLARIAH-Tech day blog

On Friday, October 6th 2017 an enthusiastic group of engineers and digital humanities scholars gathered for the third annual CLARIAH Tech Day. There was an activist mood, this time we would do things differently!

Many developers in the project wanted a meeting in which building stuff would be the focus instead of listening to presentations on how other people had built stuff. The weeks before had seen a flurry of emails on the contents of such a day and the agenda, but also on doubts and concerns. And the truth was: none of us actually had the foggiest idea of how to do this.

gertjanNo one had the foggiest idea of how to do this Tech day!

I was asked to take the lead, and together with Roeland Ordelman, Richard Zijdeman and Marieke van Erp we sat down during the CLARIN Meeting in Budapest to kick around some ideas. We settled on a hackathon/unconference-style format. The agenda would be open to suggestions from the community and not be set until the meeting itself. And I’ll confess - I had some prior hesitations on this open format: what if nobody would come up with anything? Wouldn’t people want to know what the meeting was about before making time in busy schedules? But this was what the community itself had repeatedly asked for, so damn the torpedoes - full steam ahead.

And we were not disappointed! The ideas, suggestions and questions poured in and were eventually gathered into four main topics:

  1. Integration and modelling of shared data between the various domains and the generic CLARIAH infrastructure;
  2. Continued development of GRLC;
  3. A discussion on workflows, and how tool selection based on data mime-type can provide guidance for users;
  4. TEI/exist-db/TEIPublisher and Oxygen as the basis for digital editions and linguistic querying.
techdag 1 techdag 2 techdag 3 techdag 4 techdag 5

The enthusiastic response continued into the event itself. It became immediately obvious that the restyled Tech Day would also be a lot of fun. The smiles, enthusiasm and flexibility were fantastic. The number of developers who had come from all over CLARIAH had brought many guests, turning this into a truly international day that generated a very positive vibe of its own.

After a five minute pitch for each topic, the community basically took over the pantry, restaurant and meeting rooms at the IISH building. You could find groups of engineers working, discussing and building stuff everywhere. And these groups were extremely varied: people from Media Studies discussing GRLC with engineers working in the field of Social Economic History, and Linguists and Lexicographers getting stuff done with developers working on generic infrastructure. Many new ideas were born that day.

iisgLunch at the IISH

A lot of progress was made on the four main topics. Both Open Dutch Wordnet and the first version of the diachronous lexical corpus Diamant (INT, Kathrien Depuydt and Jesse de Does) were connected to the generic infrastructure, as were catalogues provided by NISV, and the Gemeente Geschiedenis dataset on Dutch municipalities (by Hic Sunt Leones). Carlos Martinez and a group of engineers added to GRLC the automatic inclusion of SPARQL queries stored in github. And there were plenty of discussions on planned and unplanned subjects. Jan Odijk and Jesse de Does ran a very interesting meeting on workflow systems and Eduard Drenth (Fryske Akademy) presented his ideas on digital editions followed by a very detailed open discussion on the pro’s and cons of the software stack he proposed.

Completely spontaneous, Richard Zijdeman showed us a new way of implementing HDMI for the improvement of health in CLARIAH, and Roeland Ordelman and Liliana Melgar came up with very interesting ideas on a user workspace that may eventually become part of the generic infrastructure. Although interest in the first was quite short-lived, the latter we are definitely going to test.

In short: the CLARIAH tech community rallied around the open format! During the final meeting I was happy to announce that given the excitement and energy, the board had decided right then and there, that we could run another Tech-meeting in late winter, early spring 2018. And with illustrating enthusiasm the first ideas for this meeting are already coming in.

 

Gertjan Filarski