I've been wondering if we ought ask for the maven gump goal to be updated to
use <maven (instead or <ant) and/or [for short term] use both. For both, I
was thinking of something like:
<project name="X-maven ...
...
<maven ...
<project name="X ...
...
<ant ...
.. with (eventually):
1) A <depend on 'maven' for the first, and a <depend on 'ant' for the
second. We don't need the former right now, since Maven is packaged (and
also no building, so it'd be harmful).
2) With <depend (as per group in Maven) we could try with and
without -depend. The trouble is, I'm not sure this works once we start
thinking about dependencies. How would the Maven goal know what is build by
Maven, and what built by Ant -- so whether to use -maven on the end of
<depend entries. As such, I'd recommend (for now) not using the -maven in
that case.
If this change to the goal makes sense, we'd at least start being able to
generate (via Maven) good testing GOM descriptors. We could compare the <ant
and <maven results, to help us
debug. Eventually we could obsolete the <ant (especially if Maven regularly
bootstraps.)
Stefan, did we have any other feature requests for the goal? Do we need the
<mkdir entries added (so compiles pick up directory not created until part
of the build) or is this just for <ant versions, so not so important?
BTW: Brett, with Maven group/artefact dependencies and Gump <depend (group)
and id="x" (artefact id) -- but Gump having to use artefact overrides to
control Maven's environment, we are starting to depend (very heavily) on not
just group but artefact information in Maven. Are current Maven descriptors
setting explicit dependencies on all sub-project artefacts? If not ... we
have a problem.
Thanks for your attention.
regards,
Adam
--
Experience the Unwired Enterprise:
http://www.sybase.com/unwiredenterprise
Try Sybase: http://www.try.sybase.com
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
