| Re: [jakarta.ee-spec.committee] [External] : Re: Shouldn't Creation Review include an initial plan? |
Right, I thought the conclusion was to include a plan with the creation review.
If we want to allow the team time to assemble and organized,
maybe we set some kind of "initial get organized" timeframe --
maybe 3 or 6 months for the first plan review (then every 12 mos
for progress reviews)? I'd initially tried to avoid changing the
process, but it may be we need some kind of change.
Seems like we either need to adjust the description of the creation review to require a plan, or we need to adjust the diagram to insert an organizing state between the Creation Review and the Development state.
If we were to agree that this change was beneficial, we would then want to decide if this is something we inject via the JESP, or if we wanted to try to make this change in the EFSP. Probably the JESP is easier for us to change.
-- Ed
We would either want a plan with the creation review or require a plan review after creation review. The second case is what we originally did, but eliminated the plan review as redundant.
The main question comes down to when do we want the plans to be made and who do we want to be involved and have influence in that. Do we want the plans to be made by one or two people, before everyone is gathered at the table as a community. Or do we want the plans to be made once it's a full spec project and all participants are can decide together.
I'd vote the second one provides the people who join more influence.
-David
On Mar 25, 2025, at 2:54 PM, Ed Bratt via jakarta.ee-spec.committee <jakarta.ee-spec.committee@xxxxxxxxxxx> wrote:
_______________________________________________Committee members:
In the past, I think we have had some challenges with new specification project proposals when they didn't include planning details. I think we are repeating that cycle with the creation review for the Jakarta Query project.
If you recall, the EFSP process lifecycle is ambiguous on this point but some have pushed the idea that the creation review ought to have some content that identifies goals, milestones, or aspirations. Otherwise, they're going into the development phase with no stated milestones, and we with no clear understanding when a specification version might come into existence.
We are being asked to vote on creating this specification project. The best detail we have describing it is written on the project proposal page. This is a good overview, but all it says, with respect to a schedule is that work will begin this year.
But, there isn't much, if anything that suggests what kinds of milestone targets this specification might be attempting to achieve. Should a draft be expected for a milestone this calendar year? In time for EE 12? Should the working group be working to generate some resourcing for it? If nothing else, I think the review should state what the first milestone might be and, what, if anything might happen after that.
I certainly want to encourage new specifications -- but my recollection is, when we approve creation reviews, without knowing what the plan is -- when it might be completed -- in this case, when or if we might anticipate migrating and aligning requirements from one specification to another (in the case of Jakarta Query, I think the goal is to migrate requirements from Jakarta Data and Jakarta Persistence -- but, when might that migration take place and when should the impacted specifications be anticipating these changes isn't even hinted at.
So ... should a creation review include a plan and/or milestone objectives? (I think it should.)
If it should, I think the Jakarta Query proposal team should be asked to establish, at the least, preliminary goals so we can anticipate how this evolution might take place.
What do others think?
-- Ed
jakarta.ee-spec.committee mailing list
jakarta.ee-spec.committee@xxxxxxxxxxx
To unsubscribe from this list, visit https://www.eclipse.org/mailman/listinfo/jakarta.ee-spec.committee