2025-10-28 Better Sample Data Meeting notes

Status updates

• Waiting to load sample bibs in Inventory until Order app reference data is set up

• Reference data is a challenge to populate; Autumn and Shelley have plans to use .csv files and Stanford scripts

• NDA agreement with Stanford? @Lee Braginsky was going to investigate but this is moot now because an agreement has been set up and sent to EBSCO; @Shelley Doljack can check on the status at Stanford and @Yogesh Kumar will check on the EBSCO side.

1. Update for PC

• Difficulty arranging contributions of large anonymized data sets from institutions

  • Anonymization work has progressed in recent months though it is currently paused

• Difficulty understanding & populating required reference data

  • Example: Fund codes don’t work unless assigned to a budget expense class, and budget expense class has to be funded from a yearly fiscal budget

• Next steps include:

  • Explore alternatives to populating Bugfest environments if dataset contributions cannot be acquired from Stanford or MSU:

    • Solicit contributions from SIGs

    • Consider generating some random non-sensitive data where it would be useful (Charlotte mentions a similar approach was just used for NLS migration prep)

2. Outstanding tasks

• @Charlotte Whitt has invited Jason to 11/11/25 meeting, @Autumn Faulkner will send meeting link

• @Autumn Faulkner needs to supply Shelley with .csv files for reference data for Inventory data

3. Review timeline document and goals

• Review: timeline document

• Snapshot vs. Bugfest

4. Priorities for Snapshot environment

• Work with Acq SIG on reference data for Orders

• Get Inventory and Orders app populated

• What about Users app?

• Upgrade to Sunflower?

Other topics

• Bigger question for the group – who will be responsible for maintenance and updating of sample environment data long-term? Will want to recommend upon group conclusion

  • Ongoing maintenance of accessible, highly functional test environments is crucial for testing participation

  • Charlotte reminds us that the original aim was to get community testing happening even before Bugfest

  • Note: Reference and Snapshot environments are recreated from scratch each time, but same old data is re-used, not migrated up for current schema

    • Good example: Contributor type in some sample data Instances is missing, because that data point was not in the initial schema

• Do we have a useful scale of data for each type of environment and for each module?

  • I.e., 100 MARC records for Snapshot, 100 MARC authorities for Snapshot, etc.