On Tue, Apr 14, 2009 at 02:41:44PM -0400, Stevan Little wrote:
around() (without calling the next method) would also work.
That won't be detectable at composition time though.
Well, if you didn't have the method, it would fail at around() time, so it
would work in the sense that it would let
On Tue, Apr 14, 2009 at 2:41 PM, Stevan Little
stevan.lit...@iinteractive.com wrote:
On Apr 14, 2009, at 2:29 PM, Hans Dieter Pearcey wrote:
On Tue, Apr 14, 2009 at 02:22:52PM -0400, Stevan Little wrote:
So, while I am not willing to change this behavior, I am willing to add a
warning so
No, that is correct, but the actual process is that the method from
the role is composed into the class, then the around wraps it.
So no need to turn off the warning as it would not actually happen.
- Stevan
On Apr 14, 2009, at 2:47 PM, Hans Dieter Pearcey wrote:
On Tue, Apr 14, 2009 at
On Wed, Apr 15, 2009 at 12:24:36PM -0400, Stevan Little wrote:
No, that is correct, but the actual process is that the method from the
role is composed into the class, then the around wraps it.
So no need to turn off the warning as it would not actually happen.
Er, yeah, that was my point:
So, Ovid has a new Moose related post on his use.perl journal (http://use.perl.org/~Ovid/journal/38809
) which sparked a discussion on #moose IRC channel.
The core of Ovid's issue is that when a role is composed into a class,
the class wins in any method conflicts. This is how roles were
On Tue, Apr 14, 2009 at 02:22:52PM -0400, Stevan Little wrote:
So, while I am not willing to change this behavior, I am willing to add a
warning so that it is not so silent anymore. So code like this will
warn you that your overriding a class method.
This is a reasonable compromise, and
