Comet: Sprint 7 (September 14 - September 28, 2021)

What was the goal of Sprint 7?
Tying off loose ends on collections support and the default workflow as well as getting started on the batch ingest UI.

What is the milestone that this sprint is supporting?
We finally have the default workflow completely working. We can now push Objects through the workflow up to the “completed” stage.

We also have a very good start on the batch upload process. There is now a batch upload button on the taskbar on the user dashboard. This hasn’t gone through a design process yet, but the idea is that designated staff will use this form to upload a metadata manifest in a format this is supported by the system (working with CSV at the moment). We do have some infrastructure in place for parsing a CSV file and creating records from it, but we’re not actually doing that at this time. Our next step is to get to a place where a user uploads a batch and there will be some level of validation of the metadata manifest to make sure that it looks like what we want and then we’re going to create a series of records from it.

Accomplishments of Sprint 7:

• Dev side environment changes

  • The dev environment has been set up to use Docker Compose. The project team is transitioning completely to this environment.

• Batch upload UI changes

  • Batch upload button added to the taskbar.
  • Infrastructure to support parsing of metadata manifests is in place.
    • Note: Use case that we are targeting first and foremost, is that both the metadata and files will be processed together.

• Collections support

  • Can find/discover collections.
  • Can add an object to a collection (not 100% reliable at this time).
  • Can remove an object from a collection.

• Object edit

  • Logged-in users can edit metadata fields via the metadata edit form, submit, and then see that the updates have been applied.

• Default workflow is fully operational

Did we do everything we set out to accomplish?
No. We are still seeing a bit of fussy behavior with Collection addition. Also, sometimes we can add Objects to Collections, but this hasn’t been consistent. This is not surprising as this is one part of the upstream Hyrax that is the least developed for the infrastructure that we’re using for storage.

What’s next?

• Tie off remaining loose ends (collections and default workflow).
• Handle batch deposits from start to finish.
  • In addition to just having the possibility of creating batches of Objects that we can upload in large groups, what we mainly want to support is having multiple Objects within a batch go through workflow processes together.
• Sketching of UI wireframes for batch ingest workflow.

GitLab link:
https://gitlab.com/surfliner/surfliner/-/milestones/70