On Wed, 13 Feb 2002 03:07, Jose Alberto Fernandez wrote: > I need something that is to some extent compatible > with the current code in ANT1 and the way ANT1 finds > things in the only two roles it understands today: > task and datatype. The way this is done today is using > the TaskContainer interface. That is you do not look > at the parameters of the methods in the container-task > but instead you look at the interface implemented by > the container-task itself. > > This has in my opinion the advantage that the > container needs to declare what it is that it allows > to contain, as oppose to trying to discover it. > Still one can have conflicts, but it should really > reduce the chances for those.
I am not sure there is any more or less conflicts with either method but your method is more compatible with current ant model. > Proxys need to be there also (optionally) so that we > can do what TaskAdapter does AND more importantly, try > to solve the type inconsistency issue between > TaskContainers and Datatypes that now is solved by > accident. Blech - dont like adapters at all. I guess we can chalk this up to ant1.x legacy but it still feels ungainly to me :( I would prefer we just ignore the DataTypes inconsistency and just try not to introduce anymore inconsistencies. > If you mean by "mutable" the fact that a class > implementing a datatype can be changed by a > subproject, then I would have expect you to have more > confidence on my design ;-) :) No I was more thinking that you could grab a FileSet object, add a new pattern to it and this new pattern would appear in parent project. That makes me feel icky. I guess this is already possible with inheriting datatypes via ant/antcall - yuck. Hmm we haven't released that maybe we should consider zapping support for it ? :) -- Cheers, Pete The avalanche has already started. It is too late for the pebbles to vote. -Ambassador Kosh -- To unsubscribe, e-mail: <mailto:[EMAIL PROTECTED]> For additional commands, e-mail: <mailto:[EMAIL PROTECTED]>
