On Oct 5, 2006, at 7:53 PM, Aristedes Maniatis wrote:
* Andrus, would you care to give a short progress summary on 3.0?
How close is JPA to completion? It seems a lot has been done, but I
don't know others have a sense of what is left.
* What is the goal for 3.0? Is it done when JPA is finished? Are
there specific functionality goals or a time deadline goal? It
seems that many new Jira tasks are landing in 3.0 at the moment, is
it the aim that they will all be done?
JPA implementation is slowed down due to the fact that a number of
people who volunteered have other commitments and don't spend any
time on it. Sad, but this is a reality of open source. But we are
getting there... I am certainly committed to finishing it.
Making 3.0 a JPA-compliant release is certainly a goal. But I am
realistic - formal JPA compliance (meaning passing Sun TCK) can be an
insanely complex task. Right now JPA provider depends on Cayenne
core, while Cayenne core does not depend on JPA provider, so we have
a flexibility to release the core independently from JPA, and we'll
do just that if we get stuck with certification.
Regarding the core - till now we let the organic growth of features
to occur, instead of setting the raodmap in stone. Which is great IMO
(otherwise we won't get the caching features for instance - it
evolved out of a customer project). If we are to define the target
more specifically, here is my vision:
* advanced caching (already there)
* all kinds of inheritance that we need for JPA
* switch Modeler to plugin engine
* other things we need for JPA (callbacks, DataContext refactoring,
etc.)
* Does 3.0 need to be done before Cayenne becomes a full Apache
project?
I'd say after not before. In general 3.0 is not tied to the
graduation process, other than that we probably won't publish 3.0
milestones until after we graduate, just to avoid double-voting
overhead and "incubating" label on the release.
Andrus
