Technical Minutes 2020 05 08 - adewg/ICAR GitHub Wiki

ICAR Animal Data Exchange Working Group – Technical meeting 8 May 2020

  • In attendance: Andrew C, Anton H, Arjan L, Beate MF, Bert VL, Craig V, Erwin S, Kor O, Marvin R, Michael J, Mike L, Peter B, Thomas P.

Agenda:

  1. Introductions
  2. Feedback from Webinar
  3. ADE 1.0 publication
  4. Workplan for ADE 1.1
  5. Webinar participation for 9 June 2020
  6. Release notes and change log

2. Webinar feedbacks

  • There was good feedback from the main ADE working group meeting
  • Attendees found it easy to follow and a good introduction

3. Publication of version 1.0

  • ICAR can promulgate the release - they have announced this through the links to the webinar and GitHub.
  • Andrew will put out an announcement on LinkedIn and Twitter that people can share.
  • Andrew and Cesare (ICAR) created new branches - ADE-1 the default release branch, and DEVELOP for contributors.

4. Workplan for ADE 1.1

  • The group discussed the workplan for v1.1 - ie, next non-breaking changes. Agreed there should be a focus on new messages.

Priorities

  1. Body Condition Score #83
  2. Health and Treatments #96 (e.g. to support antimicrobial resistance analysis)
  3. Live animal conformation scores/inspections #97
  4. Standardise PUT/POST methods
  5. Aggregated data
  6. Group assignments (management or other groups) #88. This is different from contemporaryGroup for statistical purposes. Potentially may require a group resource or membership resources, or join/leave messages.
  • Agreed that more suggestions could be added through Issues.
  • Discussed the use of enumerated types vs. strings. Where validation is not known or essential, items have been left as strings/numbers, but enumerations are used where values would be useful.
  • Noted that medicines could be constructed with a (country or international) scheme and ID (e.g. registration within country). Some thinking would be necessary when animals are transferred between countries. The same applies to diagnosis coding.

Feedback and suggestions

  • Discussed feedback on the 1.0 API
  • Want to avoid breaking changes so early in the cycle. Breaking changes would require a new MAJOR version.
  • Most fields are optional, and developers should implement fault-tolerant reading. This way they can consume data with MINOR version additions and implementation-specific additions.

5. Webinar participation for 9 June 2020

  • Success or implementation stories (e.g. Lely)
  • Suggestions for version 1.1 and the future roadmap
  • Possibly a technical "how-to" session for developers or technical people. Would need to advertise separately. Invitations could be made directly.
  • Bert will discuss this further with Robert and other organisers.

6. Release notes

  • Suggested that we should have good release notes to track changes, for instance between 1.0 and 1.1.
  • All Pull Requests should link to the issues that they close/address (and vice versa). This makes it easy to automatically generate a change log.
  • Add to instructions for people contributing, and the checklist for those reviewing change requests.

Next meeting scheduled for 22 May 2020 at 9am CET