When reviewing a plan, the question that you need to answer is whether or not you, as a member of the specification committee, would vote to approve the eventual completion of that plan when the project returns to you for a release review.
We haven't really had to face this yet. The committee has been mostly focused on the mechanics of pushing specs that don't really introduce any new essential claims or functional changes through the process.
As we move forward, you'll have to decide whether or not, for example, a plan item is something that you would actually want to see in an implementation of the specification.
It is possible that NoSQL is further along with the specification
writing than many of the previous examples were, but the plan
review ought to include some elements of the proposed plan.
Maybe this helps?
-- Ed
On 10/9/2020 8:33 AM, Andrew Pielage
wrote:
Late to the party but adding my voice to this - I'm really
struggling to give a review of this plan as I can't find any
kind of guide or requirements for what constitutes an acceptable
plan.
As Jean-Louis pointed out, trying to use the same requirements
as a progress or release review has it fall short (e.g. EPL
rather than EFTL in TCK), and it's not really the same kind of
thing.
Is there some guidance I'm missing? Otherwise I'll have to
abstain from voting (or -1).
We were not expecting a draft spec of what already exists
in the JNoSQL implementation, but a plan to bring this to a
top level specification in Jakarta.
Basically, what do we want?
Where do we want to go with this specification?
Why?
How to get at least another implementation (otherwise,
there is no point if creating a specification right)?
We haven't got any chance to discuss this during the
meeting yesterday, so maybe we can quickly come to some
kind of agreement or at least at the same level.
The PR is using the same template as the actual
specification PRs, but NoSQL is a plan and not really a
specification.
Anyways, I tried to apply the same requirements, but
it does not sound good.
We probably need to lower the bar for plans like
this, isn't it?