ant-ivy-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Niklas Matthies <ml_ivy-u...@nmhq.net>
Subject Re: confmappingoverride
Date Fri, 28 Aug 2009 13:36:06 GMT
We'll try that, although I'm not sure that it's the right thing,
because in general we do want transitivity (i.e. when resolving conf1
I would expect to get the dependencies of other1), but when resolving
conf2 we don't even want to get the other-module artifacts of other1
(unless they are also artifacts of other2).

With regard to my first question: So it is indeed by design that when
resolving conf2 one also gets the dependencies of the mapping
conf1->other1? Because the documentation doesn't seem to say anything
about that.

-- Niklas Matthies

On Fri 2009-08-28 at 13:10h, Gilles Scokart wrote on ivy-user:
> Could you set the transitivity to false ?  Didn't it solve your problem?
> 
> Gilles Scokart
> 
> 
> 2009/8/27 Niklas Matthies <ml_ivy-user@nmhq.net>
> 
> > On Thu 2009-08-27 at 22:40h, Niklas Matthies wrote on ivy-user:
> > :
> > > In the example given on [1] where confmappingoverride is set to true
> > >
> > >   <dependency name="other-module" conf="conf1->other1;conf2->other2"
/>
> > >
> > > when resolving conf2, is the result for other-module supposed to be
> > > the configuration other2, or the union of other1 and other2?
> >
> > If it is the latter, then in the case that the existing mapping is the
> > default mapping conf1->conf1 and is not needed, as a workaround conf1
> > could be mapped to the empty set of configurations. My question: Is
> > there a designated way to specify a mapping to the empty set, e.g.
> > "conf1->" or maybe "conf1->!*", or something like "conf1->[false]*"?
> >
> > -- Niklas Matthies
> >


Mime
View raw message