Emmanuel,

thanks for this clarification.

I see one last possible issue though: I need to run the "first" build
before the "second" one, obviously.

So, I guess I can schedule them so that the second one is run a few
minutes after the first one. However, this implies that any change
performed in this time span will trigger the second build before
the first one... Is that correct ?

Or maybe, I can give then the same schedule ? But in this case,
how is the ordering decided ? Does the GROUP build has precedence ?
Is the order the same as displayed on the web page ?

--
Julien

On Wed, Jun 20, 2007 at 10:08:53AM +0200, Emmanuel Venisse wrote:
> In 1.1-alpha-2, build definition are independant and even if we don't have 
> changes since latest update, the build will run if you have changes since 
> latest execution of the current build definition.
> 
> Emmanuel
> 
> Julien Stern a écrit :
> >On Tue, Jun 19, 2007 at 08:11:59PM +0200, Emmanuel Venisse wrote:
> >
> >>You should run them in separate build definitions.
> >
> >But if I do that, the second one will never run because nothing
> >will have changed, no ?
> >And If I do a fresh build, I will loose the output from the
> >first run...
> >
> >--
> >Julien
> >
> >>Emmanuel
> >>
> >>Julien Stern a écrit :
> >>>Hi list,
> >>>
> >>>I would like to run several Maven2 goals for one project under
> >>>Continuum.
> >>>
> >>>The problem is that some of these goals should be run in --non-recursive
> >>>mode, while some others should be run as a reactor.
> >>>
> >>>One example would be to generate the dependency-convergence for
> >>>a lot of modules and then deploy the site but only for the top-level
> >>>modules (it takes days otherwise and continuum runs out of memory
> >>>anyway).
> >>>
> >>>So, what I would need is to be able to run, for instance
> >>>
> >>> mvn project-info-reports:dependency-convergence
> >>>
> >>>and then
> >>>
> >>> mvn --non-recursive site:deploy
> >>>
> >>>But of course, these should be run from the same place and the second
> >>>command should be triggered even if the first one succeeds :)
> >>>
> >>>Is there any way to do this in the 1.1-alpha-2 version of continuum?
> >>>
> >>>--
> >>>Julien
> >>>
> >>>
> >>>
> >
> >
> >
> 

Reply via email to