2021-02-01 CTWG Meeting Notes

Meeting Date

• Feb 1, 2021 

Attendees

• @Brian Dill

• @Drummond Reed

• @Rieks Joosten

• @Dan Gisolfi

• @Jim StClair

• Foteinos Mergoupis-Anagnou (from GRNET)

Main Goal of this Meeting:

To do a complete walk-through of the CTWG workflow process for one term.

Agenda 

Recording

• LINK

Presentation(s)

• None

Documents

• Specification for Creating and Using Terms

Notes

1. Update on the offline meeting on CTWG workflow (given that @Daniel Hardmanwas not able to attend today)

   1. @Rieks Joostensummarized what he and @Daniel Hardmandiscussed in their offline meeting

   2. This is documented in an update to the Specification for Creating and Using Terms

   3. The Ingest phase is as simple and broad as possible.

   4. The Curate phase is where the CTWG adds value.

   5. The Produce stage is where the outputs re formed prior to processing.

   6. Rieks and Daniel agreed that documents/inputs are owned (controlled, authoritatively updated, ...) by a single party (person or group).

   7. Rieks and Daniel also talked about versioning.

      1. GitHub has robust tools for versioning.

      2. However it could be too difficult for authors to use GitHub values. Rieks favored using human-readable version numbers.

      3. The decision was to face the problems as they would arise, because versioning may not be needed at the start.

      4. As we go forward, we could learn what's needed.

   8. @Dan Gisolfi asked about next steps, e.g., APIs for getting terms from the corpus?

      1. Rieks said that Daniel Hardman was going to do a walk-through of the complete workflow in order to determine what those APIs would be.

2. Foteinos then showed what has been developed for terminology management at GR Net (see one screen shot below)

   1. This tool, called Docusaurus Terminology, is a way for generating terminology

   2. He showed how it works based on Markdown files written in a specific format

   3. It also is able to display pop-ups for terms

   4. Once written up in Markdown, the document is submitted to Docusaurus, all instances of the term are automatically mapped

   5. Foteinos demonstrated how it worked (in earlier meetings the eSSIF-Lab framework, its glossary, etc. were presented, which show what the results of the tool can look like). 

   6. We then discussed how this tool could be used at ToIP

      1. It's currently all Node.js and Javascript, not Python

      2. It's currently a single repository for all terminology, not really designed for multiple overlapping communities

      3. These current limitations are something we need to consider closely

      4. In particular we need to look at how we can manage multiple terminologies with one overall tool

      5. We need to consider how to evolve the tooling over time

   7. Drummond asked about using this tool as a starting point

      1. Rieks suggested that @Daniel Hardmanand @Dan Gisolfishould both review this tool carefully to see how well it fits our initial requirements

      2. Rieks said that ESSIF-Labs also has several issues filed for feature requests

      3. Rieks also said we would have to open it up carefully, so as not to get swamped with support-service requests prematurely

      4. Foteinos suggested that if we have developers available, we could provide feedback for the next steps

         1. This could help drive the requirements for the next version

         2. Foteinos said the he believed that they would be open to pull requests

   8. Next steps

      1. Foteinos will check on how the tool might be licensed, open source, pull requests, etc.

      2. Foteinos and Rieks will analyze the tool in comparison to our requirements to determine what we would need in the short run

      3. Based on that analysis, the ToIP Foundation could potentially provide bounties for new features

3. Review of Decisions and Action Items

   1. @Drummond Reed to propose a separate meeting in Slack

4. Next meeting

Slides

Open

Example of Docusaurus as demonstrated by Foteinos:

Open

Decisions

Action Items