Donald,
This sounds quite interesting.  Who provides these build systems?  Would
this processing just do the compile target, or does it perform the testing
as well?  What about the databases?  Our junit bucket works with Derby, by
default.

If this type of processing could be done on Apache infrastructure servers, I
think that's great.  I was under the impression that this type of processing
was left up to those entities that were interested.  For example, Patrick
Linskey (BEA) used to do this type of processing and communicate the results
as needed.  We have picked up some of this processing for the releases that
we're interested in (IBM).  But, having a generalized solution for any
release would be nice.

Tell me more...  :-)
Thanks,
Kevin

On Fri, Jan 9, 2009 at 1:37 PM, Donald Woods <[email protected]> wrote:

> Was wondering if there would be any interest in setting up automated builds
> of trunk and the active branches, which would send out Build Success notices
> to s...@openjpa and Failure notices with context around the build failure
> to d...@openjpa?
>
> We've been using such scripts on the Apache Geronimo project and it has
> been very useful to quickly see when a committed change is causing a build
> break or unit test failures.
>
> I have a copy of our script from Geronimo updated to build OpenJPA and am
> testing it against trunk right now.  Unless someone objects or if there is
> already another build system in place, I'll create a patch that can be
> applied to sandboxes and will setup a machine I have to start building
> trunk, branches/1.2.x and branches/1.3.x.
>
>
>
> -Donald
>

Reply via email to