The scope of the proposal is a process for acceptance into an official FOLIO release.
The subgroup acknowledges that a process also needs to be defined for acceptance into the FOLIO community resources (Hosted reference envs, CI/CD, platform-complete, etc.). While related, and also important, this is viewed as out of scope for the proposal being reviewed today.
We need a volunteer for this, but have none at present.
Jakub Skoczen is working on a strawman for the JIRA workflow (closely tied to the new module technical evaluation process): began experimenting with the "Process project" workflow that ships with Jira. Will share with the sub-group this week and present conclusions to the TC on 2021-11-10.
Tod Olson: there is a nascent cross-council group working to define the larger picture of module evaluation across councils
Tod Olson: So if a dev team wants a module to be included in a new release, they need to go through the PC. Is that correct? Yes.
Adam Dickmeiss: what about adding new 3rd party-deps to an existing module? Out of scope.
Jeremy Huff: whoops, there is a discrepancy with the acceptance criteria where we specify that code must already be within github.com/folio-org. Marc Johnson: yeah, these two docs need to be reconciled.
Philip Robinson: glad to see Jira being used to manage this process.
Tod Olson: This looks great! But ... assuming we will look at it iteratively and that it may be tweaked in the future?
Yes, we should plan a retro after it's been put through its paces a few times.
This makes some assumptions about input from the PC; Tod Olson will help raise awareness of this within the PC.
Marc Johnson: the goal here is "to make a process"? Yes: the task is to specify the process.
Owen Stephens: how is evaluating a proposal a decision-making process? Proposals are approved or rejected; this is the decision. This can be made more explicit in the "Goal" statement.
Motivation: consistency! consistency! consistency! Also, to promote participation in the decision-making process.
Marc Johnson: I thought we were trying to define a framework for the process, but achieving consistency is an outcome of having such a process, so it feels like we're confusing motivation with outcome.
High barriers will limit participants; low barriers will reduce consistency; the goal is the middle ground.
Owen Stephens: Should the process favor consistency over other things, or are we expecting that the process will increase consistency? We've had presentations from e.g. sysops that demonstrated conflicting priorities among ease of maintenance, feature work, etc; are we saying that consistency is our highest priority?
Jeremy Huff: Agreed, we would have to intentionally favor consistency if we want that. Deviating from that would have to be deliberate. Should this priority be captured?
Jakub Skoczen: we've added many criteria to our acceptance criteria, but it isn't clear that the project truly will be better off if we add additional criteria, perhaps at the expense of losing a feature that was inconsistently implemented.
TODO for all: please review and comment on this document! Consider whether the RFC proposal can meet the purpose outlined here.