2025_04_30

[ChrisJohnNOAA]

Attendees:
Seth Champagne
Lynn Dewitt
Thomas Gardner
Chris John
Tylar Murray
Shane St Savage

1) Review previous action items

Shane/Chris Document setting up SSL with Docker is done (mostly Shane).

S3 work and find users currently using S3. We have several users of the current S3 integration. Seth's work got delayed due to funding, but should be coming soon.

The meeting is now set as automatically recurring. Easier to add new export formats was added to priority projects and then done.

Punted to next time, Marco: Update on datasets.xml CMS (adding here to get an update next time)

2) Go over the priority projects and discuss any changes we'd like to make.

Chris: I added some specific documentation updates (better search, migrate the information.html/documentation.html files)

Brief discussion of Java templating frameworks. Consensus seems to be there are a lot of options, and in most cases there aren’t huge differences. We should pick one and use it.
Chris likes JTE because of compile time type checking (and a performance focus).

Request for support for graph page to allow combining several layers on top of each other.
(Seth) Additional toggles for data access forms

3) Review of 2.27.0

Color bar converter by a Google Summer of Code candidate Harish
Export file type refactor
Semantic Versioning
Localized dataset metadata (in progress, hopefully close)

4) Plan what work we are targeting for the next release (2.28) and what the target release date is.

Chris: Migrate the information.html/documentation.html pages to documentation site
Prometheus configuration and more documentation to provide a basic monitoring configuration.
Google Summer of Code starting, so that will take some of my time

Shane: Erddap metrics and dashboard, sencoos might be able to give time.
Relative url is a priority for Shane, he might do it.

Shane mentioned Docker Scout, we should use it to monitor dependencies (particularly those only in the docker image- we do have GitHub Dependabot monitoring our Maven pom file).

We’re seeing regular test failures because several tests are reliant on external servers- maybe mock out calls soon. General approval for this approach.

One thing we have worked on (outside of ERDDAP) was to build AI schema.org based metadata files for data within erddap. the standard we used was called https://docs.mlcommons.org/croissant/docs/croissant-spec.html given it is an extension of schema.org it would be interesting to see if people are interested in this more
Maybe help document datasets for AI

Action Items

• (Marco) Update on datasets.xml CMS (adding here to get an update next time)
• (Seth) S3 work, checkin
• (Chris/Shane) publish ERDDAP official Docker image