Attending
Time:12:00PM PDT/03:00PM EDT - 01:00PM PDT/04:00PM EDT
Zoom: https://princeton.zoom.us/j/281265700
Participants
- Trey Pendragon (Princeton University Library)
- bess (Data Curation Experts)
- Mark Bussey (Data Curation Experts)
- Noah Botimer (University of Michigan)
- James Griffin(Princeton University Library)
Agenda
- Sprint Scheduling Doodle Poll (https://doodle.com/poll/x9ra86uts6tnbgxv)
- Assess the migrated GitHub Project
- Outline what needs to be addressed during a manual migration from Waffle
- Sprint Planning
Notes
Sprint Scheduling
Noah is not available for full-time commitment to the sprints
First 3 weeks before Summer, Mark and Bess would contribute at the capacity of one FTE (50-75% each)
From July onwards, they will be at higher capacity
Earliest date would be May 6th
Otherwise, it would be July
(All agree with this)
Trey: Proposes that the week of May 6th is the first
Tom won't be available, DCE will be at limited capacity
Week July 29th would be the second week-long sprint
Any sprints following these, regroup after May 6th
Noah: Hoping for any issues which consume significant time might be delegated to themselves
Would prefer that all participating be attentive to these situations
Assess the Migrated GitHub Project
Waffle sent Trey an e-mail requesting a seconding attempt at the migration
Noah: Exploring alternatives to Waffle.io
Bess: Finds ZenHub to have some appeal
Can have an unlimited number of projects on the same board
Trey: Princeton also evaluated this, but found Zube.io to be the preferred platform
It automatically links Pull Requests to issues, and offers label synchronization
ZenHub is free, whereas Zube does not offer a free tier of service
HubBoard should also be explored
GitHub Projects don't seem to support the linking of issues between Projects
ZenHub has good bulk management tools for manually addressing the Waffle migration
Sprint Planning
The upcoming sprints need to address the following:
- Propose deprecation of om
- Centralize on the version of Ruby and Rails versions which we test
- Centralize on using CircleCI for continuous integration
Chris Colvard has added CircleCI to a number of projects in order to support Rails 5.2 across components
With this has come the CircleCI Orb: https://github.com/samvera-labs/samvera-circleci-orb
Samvera is on a trial for CircleCI's new pricing structure
This is hourly (as opposed to the number of boxes/containers used during the testing)
We have 200 boxes and they track the amount of time which we use
If the pricing is acceptable after this trial, the enhanced performance might remain the standard
Otherwise, we might need to revisit how best to restructure build matrices (in order to avoid queuing delays)
Proposals for Other Work:
Bess: active-fedora
https://github.com/samvera/active_fedora/pull/1294
This is blocked by hydra-head (and a few other dependencies)
The value added is that the generators between Hyrax and Blacklight create different configuration directories for Solr