geronimo-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Prasad Kashyap <>
Subject Re: Migrating to maven 2
Date Wed, 15 Feb 2006 18:47:53 GMT
That's right. I have the deployment plugin in m2. I'm testing each of
the goals one by one as I use them in the itests project. The itests
project is also in m2.

I like Gianny's roadmap. We'll know the potholes and pitfalls only
when we start driving. So either we could randomly take a few modules
to convert or identify the complex ones.

Even with the complex ones, we have 2 options
1) convert them the first, they'll be the prototype for the rest.
2) convert them the last, at which point we drop all m1 and it's
baggage complete and make that final dash to m2.


On 2/15/06, David Jencks <> wrote:
> On Feb 15, 2006, at 4:49 AM, Jacek Laskowski wrote:
> > 2006/2/15, Anders Hessellund Jensen <>:
> >> We need a m2 version of the geronimo-dependency-plugin. Is anyone
> >> working on this, or perhaps it already exists somewhere?
> >
> > AFAIK, the answers are no and no, appropriately. If you'd start
> > working on it, create a JIRA task item, so it's recorded. The work
> > could be done in sandbox.
> I think we need something like plugins2 for the m2 plugins, in
> trunk.  IIUC Prasad has already converted the deployment plugin and
> it needs an accessible place to live.  I see no reason to hide these
> in the sandbox.
> Note that this plugin depends on the service-builder module.  It
> might be worthwhile rewriting the logic to avoid using xmlbeans or to
> copy the appropriate schema parts into the plugin to remove this
> dependency.  I imagine that when we have a fully working m2 build we
> can use the m2 poms instead of the geronimo-service.xml files anyway.
> I'm not sure if this idea is the same as previous suggestions, but I
> think it would be possible to stage a move to m2 as long as the
> modules built with m2 don't depend on any modules built with m1.  We
> would have a new0.5 goal that would build the m2 modules, and the m1
> goal could have the m2 modules excluded in the call to the reactor.
> thanks
> david jencks
> >
> >> /Anders
> >
> > Jacek
> >
> > --
> > Jacek Laskowski
> >

View raw message