Are there any plans to introduce something like contextual mixins? that is, some way to decide which mixins to apply depending on context (not in a layer/module sense but rather in a compositional context).
Lets assume for the sake of the argument that we have two different types of "Shapes" one "Polygon" based (e.g. rectangles etc) and another "Line" based (e.g. lines, splines etc) shape. Lets also assume that we have an "Selectable" fragment. If I create a composite llike "interface Spline extends LineShape , Selectable" then the "Selectable" fragment could imply the use of a "LineShapeSelectionRenderer" While "interface Rectangle extends PolygonShape , Selectable" then the "Selectable" fragment could imply the use of a "PolyhonShapeSelectionRenderer" I'm well aware that I could decorate the Spline or Rectangle with different mixin annotations, but that would require me to both apply annotations and extend the interface from Selectable I'm also aware that I can create two different sub interfaces, eg SelectablePolygon and SelectableLine which are annotated with different mixins. Both of these approaches are IMO too detailed. I would like to just say "apply this behavior/ability" e.g. "I want this shape to be selectable" not "I want this shape to be a selectable polygon" Otherwise it feels like I'm starting to fit things into an inheritance hierarchy again just like OOP. //Roger
_______________________________________________ qi4j-dev mailing list [email protected] http://lists.ops4j.org/mailman/listinfo/qi4j-dev
