> > >> Of course, if people think that this hasn't been debated enough then we >> can continue the conversation and have the vote later. >> > > Let's please consider the fact, as pointed out by Antoine, that there is > more or less no way to prevent manipulation of references by Java or > scripted tasks. That being the case it doesn't seem like a good investment > of cycles to prevent this task from functioning. With my just-committed > bugfix to RuntimeConfigurable, my current cut of AugmentReference code is > completely self-contained which means that any third party could write the > task and our proposed 'final' attribute would be meaningless. What might be > doable is creating another task that would make read-only a reference by > dynamically generating proxies that would hide setter-type (any > parameterized?) methods, or something along those lines. _That_ could stop > <augment> and other avenues from modifying a reference if that's really what > we wanted. So I'd say these are separate concerns. > Sounds good to me
-- Jean Louis Boudart Independent consultant Project Lead http://www.easyant.org
