On 17/03/07, Steffen Mueller <o6hhmk002@sneakemail.com> wrote: > I created a patch (including tests!) to make AutoLoader work with > class inheritance: so now you can have a base class and a derived > class that both use AutoLoader for some of their methods, and calling > an auto-loadable method in the base class on an object which isa > derived class will no longer result in an error, but the base class'es > method will be loaded and executed. The can() should now work as well > correctly. I haven't really the energy to re-read the patch right now, but the question that pops up into my mind is -- does this goes against documented behaviour and/or a core regression test ?Thread Previous | Thread Next