RE: cvs commit: ant/src/testcases/org/apache/tools/ant/taskdefs P reSetDefTest.java

[Dominique Devienne <DD...@lgc.com>]

> From: Peter Reilly [mailto:peter.reilly@corvil.com]
> >Is the magic ant-type attribute described anywhere Peter?
> No, Well it is on the dev-ant mailing list ~july 2003.
> 
> http://marc.theaimsgroup.com/?l=ant-dev&m=105846981518008&w=2

I remember it was mentioned, but I forgot about it until your commit.

> >It obviously allows polymorphism without using id/refid,
> Yes
> 
> >but when/how can it be used? Is it part of 1.6? Thanks, --DD
> Any addX(Y) or addConfiguredX(Y) but not createX() method.

That's logical. I always thought the createX() methods should have
been used as factory methods only, for the purpose of derived-tasks,
and the created instance would then be passed to the addX() method...

Currently, a derived task as no way of forcing the framework to
instantiate a derived-type to pass to an addX() method. You've
added a way the build writer can use, not the task writer.

The createX() allowed the programmatic polymorphism I'm talking
about OTOH.

> It is part of 1.6, but it is undocumented. (I am not too sure it is
> a good idea..., too many createX()s are out there).

I can see how polymorphism will never work everywhere because of BC
issues... --DD

---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@ant.apache.org
For additional commands, e-mail: dev-help@ant.apache.org