[bv-dev] Cross parameter validation convergence

[Emmanuel Bernard]

On Thu 2012-09-06 11:20, Matt Benson wrote:
> On Thu, Sep 6, 2012 at 9:38 AM, Emmanuel Bernard <emmanuel at hibernate.org> wrote:
> > On Jeu 2012-09-06 11:29, Michael Nascimento wrote:
> >> On Thu, Sep 6, 2012 at 11:13 AM, Matt Benson <mbenson at apache.org> wrote:
> >> > My inclination here would be to require unique signatures, and apply
> >> > full JLS method matching rules.
> >>
> >> I agree, they have given plenty of thought about it while working on
> >> the JLS and I don't think we can do much better - but there is a lot
> >> of room for doing worse.
> >
> > The rules are quite complex so before we end up having coding them, I
> > want to make sure every one likes them more than direct matching :)
> 
> I could live with direct matching, at that.  In fact, that might
> actually be better in this case.  Mainly unique signatures is the
> important thing, and we'll already avoid collision between a type-safe
> validation method and the generic method by hosting them on different
> interfaces, right?

Yes, that's the direction we are heading it seems (re hosting).