2:40 a.m.
Hey,
Some pointers for those who haven't worked on the OGM code base in a while:
* Contribution guide: http://hibernate.org/ogm/contribute/
* Build instructions: in the readme.md in the root dir;
* You need to install CouchDB separately and configure the COUCHDB_HOSTNAME
env variable if you want to test that backend; By default only the code of
this module is compiled
* When running the CLI build, MongoDB is started and stopped automatically
via Maven. For running a test against MongoDB from within you IDE, MongoDB
needs to be installed and started separately.
* There are two types of tests in core: a) unit tests for stuff in core
itself and b) the "backendtck" which are high-level (i.e. Session/EM-level)
tests and which are executed for all backends. To run a specific test
against a specific store in your IDE, simplest is to copy it into that
project and run it from there.
* Packages are organized with the SPI/impl split at the lowest level (I was
talked into that ;). SPI is geared towards grid dialect implementors. The
split is not complete yet, at some parts "our" dialects refer to
"impl"
classes from core. Anything not "spi" or "impl" is public API.
Naturally,
that's not very much though as OGM's main user API is JPA/Hibernate ORM.
The main thing is the configuration API.
Let me know in case there is anything I can help with to get you started.
--Gunnar
2015-02-23 9:15 GMT+01:00 Gunnar Morling <gunnar(a)hibernate.org>:
Hi,
Thanks for grooming the backlog and preparing these releases!
I'm curious how the weekly release scheme is going to work out. In the
last two sprints we created none (although I thought we wanted to release
HS?) and one (HV, as planned) release. Planning for four releases now may
be over-stretching it a bit, esp. when splitting up powers as we do in the
first week. I would have planned for 4.1.2 and a 4.2 release.
As you say, it's too many issues, we will not be able to do them all.
Personally, I like a realistic schedule more, IMO it gives you a better
sense of achievement if you actually managed to tackle what you had
planned. Now we'll more likely get the feeling that we failed to do half of
the things planned (as we know that's expected, still I find it
sub-optimal).
The 4.1.2/4.2 release split make sense. It may be a good idea to either
pull you, Emmanuel, or Sanne (which I wouldn't count as a newbie in terms
of OGM) over to 4.2 issues at some stage, depending on the progress we make
with the respective backlogs.
Nice to have the five of us working on OGM, looking forward to get new
functionality out there!
--Gunnar
2015-02-20 11:07 GMT+01:00 Emmanuel Bernard <emmanuel(a)hibernate.org>:
> I have walked through all of the issues and sorted what I think should
> be done.
>
> ## Plan of action
>
> Within these 3 weeks, we will do:
>
> * 4.1.2 (bug fixes) -> code freeze Friday 27th
> * 4.2 Alpha1 -> code freeze Friday 27th
> * 4.2 Alpha2 or Beta we will see -> code freeze Friday 6th
> * 4.2 Beta or CR we will see -> code freeze Wed 11th
>
> The rookie team + me will work on the 4.1.2 tagged issues.
> This work will continue even past the release of 4.1.2 but will be
> likely incorporated into the 4.2 branch.
>
> Davide and Gunnar will work on the 4.2 tagged issues.
> It might make sense to have a single one focus on the numerous query
> improvements and the other on the rest. See below.
>
> Take the top issue of the list ordered by decreasing priority or one
> with the same level of priority. Kill it. Move to the next.
>
> ## Priorities
>
> For 4.1.2, I have sorted issues by decreasing importance with one little
> twist. All issues created after a direct user feedback is marked
> blocker. This is to speed our feedback loop time.
>
>
>
https://hibernate.atlassian.net/browse/OGM-740?jql=project%20%3D%20OGM%20...
>
> For 4.2, it is also sorted by decreasing importance but I also gave the
> release a theme.
> The main focus is significantly improving JP-QL queries.
> Secondary and must have target is the error report API.
> Tertiary is everything else. That includes some improvements in ORM
> itself to unlock us.
>
> There are too many tasks to address them all but let's first focus on
> making a dent up to the major priority level.
>
> Remarks?
>
> Emmanuel
> _______________________________________________
> hibernate-dev mailing list
> hibernate-dev(a)lists.jboss.org
> https://lists.jboss.org/mailman/listinfo/hibernate-dev
>